This designation is usually reserved for foreign adversaries/companies, and so this is crazy to apply it to US company over a sudden contract dispute... that was previously agreed upon by all parties.
This should make any US company nervous about entering into an agreement with the government. Or any US company that already has a contract with the government. If they one day decide they don't like that contract, they can designate you a supply chain risk.
Not 1) rip up the existing contract and cease the agreement or 2) continue (but not renew) the existing contract or 3) renegotiate terms upon renewal but instead a full on ban of doing any business with an entire industry/sector.
Shame you didn’t donate $25 million to Trump, like the company we decided to give the contract to instead did, who will benefit tremendously from you being designated a supply chain risk. Maybe next time you’ll be a little smarter.
Well. Unsurprisingly fascists will do a fascism – an ideology somewhat defined by merging state and industrial powers. Many economically minded people, many technologists, including in this space, have afforded themselves the luxury of not talking about politics too deeply. As I said some years ago: Ignoring politics has its way of coming back to haunt you. Back then this was an unpopular take.
I think we should really judge governments by their actions and stop labeling them democracies, if they do such things that don't look like democracies at all.
I think a good play by the Democrats would be to say that if they get into office they're going to investigate all these deals as potential bribery, fraud and corruption and that any business leaders that appeared to benefit from contributions might be prosecuted. That would be a laugh, I'd love to see how quickly the excuses start rolling in.
I love when a Republican does something awful the response is "but what about if Democrats do that same awful thing to us!" as opposed to discussing and admitting that the Republicans did something awful.
How I wish this was true... Every single time we experience something (and of course lately it feels like a daily experience) I would be in a discussion at some point with a Republican and would come with super-solid counter-examples like "imagine 2029 and President AOC doing ____" - it just never works...
The democrats don't tend to abuse power back. The SCOTUS ruled presidents have immunity for prosecution for official acts, while Biden was president. He did nothing with the power.
I legitimately wish the Democrats were half as radical as Republican propaganda makes them out to be, then at least we might get something out of it before the inevitable right-wing purge.
Oh. Before your comment I completely misunderstood "Democratic administrations". I understood it to mean administrations of countries that are democratic, not an US administration that is dominated by the Democratic party.
I think you're misinterpreting the discussion here. Democrats are precommitting that they are going to do the same awful thing; when the time comes, I will be contacting my legislators demanding that they do to OpenAI or SpaceX whatever is done to Anthropic now. It's outrageous that Sam Altman would step in to try and benefit from the political persecution of his main competitor and we must ensure that he regrets this.
Yeah, now that this door is cracked open, it's now possible to decapitate SpaceX, which is at least as natsec-critical as Anthropic. The owner is a drug addict, has business interests in China, and is a Russian sympathizer (recall all the restrictions on Ukraine using Starlink), which all together is way stronger evidence for SCR designation than anything Anthropic has done. They're quickly going to come to regret opening this can of worms, but what else is new.
Trump isn’t planning on ever leaving office before his death. His sycophants will just say yes in the hopes of unconditional pardons. They know they’ll never hold a position of power again so they’re grabbing everything they can while they can.
> That’s ultimately why Ted Cruz spoke out about the Kimmel cancelation. It doesn’t take long until those powers are turned against you.
Meh, I think it's entirely asymmetrical in this era. Democrats aren't good for much, but they're very good at respecting norms.
Trump is willing to do completely unprecedented, vindictive, and malicious things because he's so popular with so many people who are either checked out, nihilistic, corrupt, or just completely unconcerned about the concept of good governance.
It's not a pendulum where there's some super-corrupt Democrat waiting in the wings to do the same things upon their enemies, this really is the Republican party openly embracing kleptocracy and lawlessness.
Like gerrymandering, I have the strong suspicion that Trump voters won't be incentivised to vote for norm respecting leaders until a Dem does very Trumpian things to their side. I'm thinking firm 2nd Amendment reform enforcement with a rapidly resourced federal agency. Then, standards will be rapidly rediscovered.
What people seem to refuse to accept is that democrats won't have another chance, any time soon. It's done and gone. Count one or two generations, at a minimum, under the new Epstein class regime, before people may try to rise.
If democracy in the US falls apart, it will take an event of World War 3 scale to fix this.
But thinking of it, your estimate might still hold under that premise.
As a European, I have still hopes for the US though. The western world is already reorganizing itself without the US in the center, but we and the world in general still need the US as being more of an ally than an adversary (at the moment, it's more of an adversary to the western world, sadly).
“Concerned” is an understatement. USA is already operating at nazi Germany levels and more than half of the civil society is approving. Not that it’s a surprise for global spectators though - it’s finally showing it’s true colors.
Creating a private militia, silencing dissent, declaring wars without congress vote… I don’t see how this is being allowed to happen without public approval, or at least, public apathy.
If civil society is not concerned by the tribute-based Board of Peace that gladly invited Putin, an attack on another country without authorization by Congress, a threat to seize the land of an ally, and the killing of its own citizens by the ICE militia, then an unjustified supply-chain risk label won't cut it.
I don't know. There's a certain segment of "civil society" that's pretty much OK with anything as long as it doesn't threaten the Holy Free Market. Free only for appropriately holy values of "free", mind you...
Nope, they don't have that capacity. It's been shown multiple times in the past year.
Shutting down USAID being the clearest one. They just saw "they help brown people in other countries with our money" and shut it down. Fuck all second and third order effects that actually benefited the US.
Anthropic should never have gotten into bed with the military or intelligence services to begin with. They wanted to make a deal with the devil and dictate the terms, that is the problem. If they had stayed out this wouldn’t be happening. Yes, someone else will probably step in and do all the evil you have just refused to do, but that isn’t a reason to instead decide to do it personally.
Note that I give them a lot of credit for trying to stop and to have their own red lines about the use of their technology, and to stick to those red lines to the end.
According to legend the devil adheres precisely to the terms of the contacts he signs; it's usually the foolhardy peasant who didn't notice the fine print.
I was sympathetic to this line of reasoning but I feel it's repeatedly shown to be self-defeating.
What chance have the proverbial good-guys got if, even after _proving_ some modicum of good will, people will nonetheless condescend any attempt to influence bad/wildcard actors? It feels great to tell someone they 'should've known better' but I'm convinced that that's basically void of cautionary utility.
I've commented this in a different thread, but pretty sure something very similar would've happened in they refused to "get into bed with the military or intelligence services to begin with".
It's damned if you do and damned if you don't — lose-lose scenario either way.
The military is perhaps the biggest possible customer around. They do plenty of things that aren't blowing people up. It's not bad to help with non combat tasks.
Congress must approve any renaming the Department of Defense... They haven't. Stop giving them what they want without them even doing it in good faith at all.
Pretty ironic because for once, they went with the more accurate description instead of the euphemisms. It's no longer "protecting freedom of speech" or "ending wars," just full-on bombings and invasions with glee.
They'll do nothing. It's really hard to take the morals of these devs seriously when they're already fine working for, and have a history with, some of the most evil companies in current existence.
And when you suggest unions and worker organization to them they gloat that the company takes better care of them without it. They don't want to be in the same class as the peasantry.
It was really easy to close out my ChatGPT account and switch to Claude. I was really only there out of inertia. I don’t do anything beyond occasional free tier stuff like rubber ducking but so far Claude is so much better.
I prefer the claude code CLI interface for everything anyway. It is actually more convenient. Memory is local files, type one word to use rather than navigate.
But for how long? If Claude is a supply chain risk. That means anyone hosting him would also be a supply chain risk.
Ergo AWS/Azure/GCP - nobody will host them because it’s Anthropic or the lucrative government contracts. Hegseth/Trump didn’t just say “you’ll never do business with the US” - it’s that they will never do business IN the US. Hopefully that means they’ll be able to take up shop elsewhere in the world.
What's getting lost in the politics here is the technical question: how do you actually enforce a usage restriction on an AI model?
Anthropic's redlines - no autonomous weapons, no mass surveillance - are policy statements. But once a model is deployed inside a customer's infrastructure, those redlines are essentially honor-system. There's no runtime enforcement. No cryptographic proof that a human authorized a specific action. No audit trail that connects a model output to an approval decision.
This is the fundamental gap. You can write all the acceptable use policies you want, but without an authorization layer that sits between "model produced output" and "system took action," those policies are unenforceable at the technical level.
The military already solved a version of this problem decades ago with chain-of-command and rules of engagement. The missing piece for AI systems is the equivalent infrastructure - a way to verify that every consequential action has a signed, auditable authorization before execution.
Whether you agree with Anthropic's position or not, the fact that the entire dispute comes down to "we don't want our model used for X" with no technical mechanism to enforce that should concern everyone building with these systems.
I always find it interesting to listen US citizen answer "What would it take for you to not consider your country a democracy?" and admire the wide range of answers and denials.
If you're referring to ICE, that's gross hyperbole, and honestly a little insulting to people who live / have lived in regimes with an actual secret police.
The US is still a rights-based state, which means that when they arrest someone (legitimately or not), lawyers and human right advocates can eventually track them down.
When a secret police disappears someone, they actually disappear. Families can spend years wondering if their loved one is still alive, or was murdered by organized crime, or ran away, or was secretly taken by the state. The US these days is pretty bad, but it's nowhere near that bad.
The arrest of a Turkish graduate student in Boston looked a lot like a kidnapping.[1] More recently, ICE responded to a judge's order to release a detained refugee by threatening to detain her family and send them to Texas if they came to pick her up and then forced the minor child to stay in a hotel room with three agents. [2, p.8] These may not be cases where people are secretly being taken by the state, but it's not hard to see why people might call the government organization detaining people and moving them around so as not to be found "secret police" or "disappearings".
Have you been paying attention? There are reports about hundreds of people going missing in ICE detention. Maybe they aren't being shoveled into mass graves, but if we don't where they are and can't reach them and ICE themselves don't know where they are because they no longer exist in their databases, is there any difference for their families?
These bullies wilt when everyone stands up in one voice. But when some parties capitulate (OpenAI), it sets a precedent that this behavior is OK. And then it’s not long until you become the target.
Since the end of WW2, and especially since the end of the Cold War, Democratic administrations have presided over significantly higher job growth than Republican administrations.
On the contrary, hopefully this gives the next democratic administration ammunition to take down big tech. Might as well classify Meta, Microsoft, Amazon, and Apple as supply-chain risks too with this logic.
Too bad that Congress has abdicated their responsibility to the executive branch, no reason why Congress couldn't have more control over the Pentagon. The President only has legal authority to command forces, not control an entire institution; but this would require Congress actually doing their job and not justifying more corporate welfare forever.
A lot of people love watching dumb shit, like watching reality tv. Crucially also they often love doing dumb shit. It's a privilege they don't want to give up, like pretending gravity doesn't apply to you.
Exported all my chats and deleted my ChatGPT account yesterday. The current administration not liking you is the strongest signal I could possibly have to go all in on a particular company.
I canceled my subscription, but have not yet exported and deleted because I'm an idiot, and also because I'm not sure if deleting it will have any actual impact (is it a hard delete? Likely not, even if they say it is).
And I'm just trying to play out what happens if Anthropic, and Google (if they haven't already), capitulate. Am I just going to forego using the best models and suffer any repercussions of not having access when the people who couldn't care less if the military is using AI for illegal uses continue to leverage them? When I say illegal I'm talking about the surveillance-of-US-citizens red line Anthropic would not agree to. The autonomous weapon one I'm sure there are zero laws against and so that wouldn't actually be illegal.
Even if you don't like the current administration, the rank and file are still out there doing valuable work. The government is more than ICE; it also administers welfare, funds research, collects taxes, and distributes social security payments to the old and infirm.
Don't get too attached. We're witnessing capitalism in its most ruthless form. Any of these companies will discard their principles the moment it becomes existential.
Losing a contract with the Pentagon and potentially all Federal-interacting businesses sounds like a pretty severe monetary hit. One which is hard to recoup by a bunch of $20/month consumer subscriptions.
I said it before and I say it again: If openly bribing a crony gov to cancel your competitor is now the de-facto standard of making business in the US, I don't see how any rational investor could still see US companies as a secure investment. When the rule of law degrades into pay-to-play politics, the inevitable result is a mass exodus of both capital and top-tier talent.
And to add to this quoting another commentator on the issue: First the Meritocracy goes, then the Freedom goes.
> I don't see how any rational investor could still see US companies as a secure investment
You are right it won’t be as secured as before but it’s only risk management. As much as investing in a oil company in Brazil is a risk because you could have their government takeover the company to make it part of the government and screw you in the process.
Rational investors live in reality. In reality, a great deal of business conducted throughout the world involves graft; companies accept that, and keep doing business.
It's not a good thing, AT ALL. There's a huge loss of overall productivity when you have corrupt systems (see Russia), which is why modern governments have worked so hard to lower corruption. But Trump ruining all that isn't going to end business ... it's just going to make everyone pay more for everything.
> which is why modern governments have worked so hard to lower corruption
I would argue that they did not. They should have and some were better then others.
But, bulk of financial markets, all of predictionmarkets and crypto, startups and sillicon valley, Musk imperium, Thiel, Murdock, all run on corruption. And to large extend, Trump is the endgame of that.
There is a substantial difference between the standard lobbying and greasing the legislative wheels, and what's going on with this current administration.
Even if companies were pretending to play by the rules before, at least they had some need to put in the effort to pretend. When a society can see belligerent ostentatious corruption going on as the norm, nothing good can follow.
That is already started to happen but it cannot happen overnight. Not only is it not easy but finding alternatives is also not easy. Just think of from your own personal perspective, say you have $100m right now invested in US business and wisely you say "I gotta get my shit away from this mess" - where exactly would you park your assets? You will find a way of course but you won't be moving $100m elsewhere overnight
> I said it before and I say it again: If openly bribing a crony gov to cancel your competitor is now the de-facto standard of making business in the US, I don't see how any rational investor could still see US companies as a secure investment.
Arguably large parts of the market in the US have been irrational and largely vibes based for a long time at this point. This action (like many others coming out of the Trump administration) adds to the chaos but I tend to doubt it will be the event that causes Wile E. Coyote to look down.
The consequence is that any company that does business with the U.S. military, and potentially any company that does business with the government in general, must stop using Anthropic's products for that work.
Anthropic has vowed to fight this designation in court.
Without weighing in on the constitutionality or legality of the move, I think it's obvious that this kind of retaliation power is unmatched by any private business that has a contractual dispute.
If a private business doesn't like Anthropic's terms, it can walk away from the deal, but it can't conduct coordinated retaliation with other companies before ending up in antitrust territory or potentially violating the Sherman Act.
Now for my editorializing: The fact that Pete Hegseth is willing to apply this type of designation against a U.S. company simply because he doesn't like its terms is pretty chilling. It's all the more scary once you consider which terms he objects to.
FedRAMP and FedRAMP adjacent revenue is non-negotiable for vast swathes of businesses. The designation of "supply chain risk" is viral in nature because no GRC team will dare take such a risk within their supply chain because most customers add BOM requirements in contracts so this will end up falling under those already.
There's a lot of backchanneling going on between Emil and Dario because everyone's in the same circles but it's all for naught.
The DoD has been rather consistent that they will decide what to do with a product sold to them, not some random vendor. There is nothing extra to "price in".
The "extra" is that the government is now attempting to unilaterally renegotiate contracts, and if the contractor disagrees, not only do they terminate the agreement but they restrict how other companies can work with you.
If you were a contractor to DoD (no way I'm calling them DoW) would you take the risk of doing business with a company that has been labeled a supply chain risk by your main customer?
The issue is the onus is on the contractor to prove that Anthropic technology has not tainted US government contracted projects - this is a herculean task verging on impossible. Additionally, most contracts will mandate SLAs around removing BOM risks.
I’d like a lawyer to give some input. If you have a company that deals with the military does this chain down to not being allowed to use Claude or not?
IANAL and this is my understanding of the situation (I can be completely wrong) but yes, any company that deals with military cant use Claude (anthropic)
In fact adding onto it, IIRC this is the reason why google and amazon have to divest essentially from Anthropic if they want Govt. contracts
Hope this helps though a lawyer's input will definitely be more credible. So its good for them to respond as well.
I can be wrong but I looked up online and there are sources which say the same, so I am curious to know more about it.
> Former Trump Al advisor Dean Ball has warned that the White House's decision to designate Anthropic a "supply chain risk" could blow back on some of the biggest names in tech-including Google, Amazon, and Microsoft-all of whom have billions of dollars riding on the Al startup. Ball, reacting to Defence Secretary Pete Hegseth's order barring any military contractor or supplier from doing business with Anthropic, said the move amounts to "attempted corporate murder."
> He added that if Hegseth gets his way, Google, Amazon, and Nvidia would effectively have to divest from Anthropic-a company they've collectively poured billions into.[0]
> Amazon alone has committed up to $8 billion in Anthropic. Google has invested around $2 billion. And Microsoft, while not a direct investor, relies on Anthropic's models through its Azure cloud platform. A forced divestiture or business cutoff from any of these companies would send shockwaves through the AI investment landscape at a time when hundreds of billions are flowing into the sector.
Writing out a thought I had, someone please critique my reasoning here...
What if Anthropic just shrugged, dissolved the company and open-sourced all of the Opus weights? Could this harm OpenAI and advance AI in a reasonable way?
Look I know it's an insane idea. I'm just curious what the most unhinged response to this might be.
I kinda wonder if this is how we got DeepSeek. It was developed by a Chinese hedge fund. Entirely possible their business model was to take out large leveraged puts against the major U.S. AI vendors; shit on their business models with an entirely open-source model; and profit. The stock market certainly dropped in a massive way when DeepSeek was released, so if they traded against NVDA/GOOG/META et al, they profited in a big way.
They would never do this because the entire point of the company is to try and control what AI is allowed to do, who is allowed to use it, and what they’re allowed to do with it. The overarching philosophy of Anthropic is explicitly opposed to open models. If it were up to them it would be illegal to inference them in the U.S.
If I were to download those weights I can't run them unless I spend $100k on a cluster, so the privacy advantage is not there yet.
We already have Groq, Celebras and AWS Bedrock and others in the inference of open models space, so the model is usable that way.
Is Claude better than Llama, Gwen etc. Probably. For now.
But for how long? Dissolving means relying on Meta or Deepseek etc. to pick up and carry on tuning. Otherwise it'll be as useful as a GPT2 or Atari ST eventually in a competitive environment.
Also open sourcing the weights is handing it over to DoD (aka DoW).
Complicated question but probably not the best move. Keep going means keep working on safety research.
> Look I know it's an insane idea. I'm just curious what the most unhinged response to this might be.
I mean what if all the employees stripped off their clothes and walked through the streets naked while barking, then called up their middle school math teachers and barked live dogs then moved to a commune and stood on their heads.
> Writing out a thought I had, someone please critique my reasoning here...
I mean to critique your reasoning, it makes sense to also include a criteria of something they might reasonably do. There are an infinite number of unhinged things a group of people could in theory do. But maybe start with something they would actually have an incentive to do.
Why would they voluntarily dissolve their company, put themselves out of work, release their crown jewels and get nothing for it? Yes it's unhinged but unless I'm missing something bug, they wouldn't do that because they wouldn't at all want that to happen.
Labeling Anthropic a supply chain risk only because they were uninterested in doing business with the US government under the terms requested seems very much a bullying tactic that results in something the west critiques China for: coerced alignment.
Wonder how long it will take the American public to designate the US Govt a threat to national security, and using AI to assemble their own autonomous civilian defense robots to protect the public from the government-approved population suppression robots.
This is also the part i don't get. People flocking over to pay anthropic which _has already been used in a war_ and cancelling existing subs to a provider that has not yet but will?
Ethical boundaries seem difficult to draw here. I don't really see people taking the stance of "No longer paying any of them" which would make a bit more sense to me.
Anthropic already had layed in bed with pentagon, how did that fit their overall ethical standpoint as they were already being used and before they tried to walk back their terms?
I don't get the impression Anthropic tried to walk anything back, they just balked when they were informed that the DoD wasn't willing to abide by their terms to refrain from using Claude for surveillance or straight up selecting targets to kill. But you're right that means Anthropic's hands are not exactly clean, just less bloody than OpenAI's or xAI's
Is there a link to the actual order anywhere? For us FedRAMP folks, the exact order contents actually matter, rather than a journalistic regurgitation. I was hoping one of the links in the article pointed to a source, but they're all just links back to other WSJ pages.
Most corps will take the safer bet of not using them at all to prevent "accidental" miscompliance and more expensive audits. I'd bet if you need fedramp compliance you'll get forced out of their products generically.
It sounds like they still have not issued any sort of actual order. The "formal label" described in the article is that they sent a communication directly to Anthropic saying they're a supply chain risk.
I think the DNC and the media might need to get some of that blame for being empty vessels for corporate interests that allowed this conman to get elected twice
Voters knew who trump was and chose him. They deserve all the blame. As a voting adult ur choices have consequences. All voters who voted trump or third party deserve all the consequences
Of course. Hegseth said it, there is no way they could back out. Looking 'weak' is the worst possible thing for this admin. They would rather look childish, stupid, and evil, as long as they don't look 'weak'.
Especially 'weak' things like 'caring about people'.
> For what it's worth govts should have the final say over companies they do business with
That's certainly an opinion, but in the US it's not how it works. Doing business with the government does not give the government total power over the company, that would be absurd.
US rich tech folks like Dario only get a spine when there is money on the line. Where was he when US govt was doing awful illegal stuff against non tech companies and other americans... All of this is practically just theater at this point.
From what I understand it cannot be used to perform work on contracts where the DoW is on the other side. [1]
In practice I would suspect companies with such contracts would play it safe by outright banning the use of Anthtropic products, even if they could technically be used for work on contracts with other parties.
Anthropic can now no longer buy new hardware and probably will be kicked out of all cloud compute. They can not also move to a different jurisdiction as exporting model weights is now considered same as exporting ICBM technology. Wow, companies in China are now more free than Anthropic. It's a death sentence, and a huge win for OpenAI.
> Anthropic can now no longer buy new hardware and probably will be kicked out of all cloud compute.
That's beyond the authority of the DOD. What the DOD can do is say that Anthropic's products cannot be used on defense contracts. So Boeing programmers can't use Claude on a defense contract, but they can still use it on a civilian contract. Similarly, AWS or Azure can still be used by Anthropic, but they cannot use Anthropic in their defense work.
As much as I'd really like to see one of these labs in EU, they are likely way too tied to US capital and supply-chain of semis/services. If they move away and now have to also deal with export restrictions they may make their existence harder.
I hope so. I will never type a single thought of my own or personal detail into an OpenAI product again. I have no doubt at some point OpenAI will be asked by DoD to hand over customer data and they will do so. If I use AI at all for nonprofessional reasons it will be Anthropic/Claude.
US Gov doesn't really need a contract with anthropic to force them to hand over customer data do they? What would prevent that from happening if they are still a US based company?
I am a political moderate who dislikes both the Democrats and Republicans. I think that I have been fair to the Trump administration in the past, including occasionally defending them from some of the less reality-based accusations against them.
I canceled my ChatGPT subscription a couple of days ago. In my opinion the Trump administration has become far too much of an "imperial Presidency" in its acts of war and its attempts to bully companies. It is also corrupt on a massive scale. I distrust anyone who thinks "yes, I'd like to work with this administration".
Could this be the chain of events that finally pops the AI bubble? If OpenAI's reputation hit slows growth enough to scare off investors and Anthropic's growth stalls due to this government attack...
Of course it's going to be this government that goes and pokes the bubble that's propping up the economy, despite all the government's other shenanigans.
Anthropic was very clear about the usage restrictions: They didn't want them being used to control autonomous kill drones or mass surveillance of the American public. That's it. DoW didn't like that -- for reasons that will probably soon become apparent.
Correct, it will be about silencing any opposition against this administration. OpenAI will be happy to let their models be used to persecute, kill, and destroy american democracy if it lines Sam's pockets.
either you allow a democratically elected government to do everything they want that is legal, or you insert private corporate decision-making into every government decision which is untenable
Is there any evidence that going outside the scope of the agreement would amount to anything more than a contract violation? Are we really to expect that Anthropic general counsel sits at the API gates allowing or blocking requests?
More generally, are there any comparable contract requirements in the field of defense, for a company in the same position as Anthropic? I'm curious.
You're missing the huge step that the government asking for "all legal uses" terminology is also who decides what is legal. Congress isn't willing to act as a check on executive power, meaning the contract they demanded simply says "I do what I want."
Once again our leadership is "playing government" like a bunch of 12-year-olds, lashing out impulsively without thinking of the consequences. And no doubt once again it'll take a year for this to wind its way through the legal system and be reversed long after the damage is done, as is finally happening with the tariff fiasco.
Conservatives haven't had any power in Washington in decades. They are in thrall to MAGA now, which is all about seizing the means of production when its convenient.
Hate to break it to you, but Trump and MAGA are conservatives in the US. The conservative and Republican brand is tied to that wagon now. They're going to have to make a serious effort to separate the two if they ever want to get rid of the Trump stink.
As the sibling explained, it was sarcasm. It was satire of brain dead right wingers saying "stop making everything political" when political points apply, but then making things political politcal in stupid situations.
Example, someone shoots up a school, people say "if the Republican party hadn't obstructed gun legislation maybe this wouldn't have happened", they respond "StOp MaKiNg EvErYtHiNg PolItIcaL". But if the person is trans suddenly it's political.
First, I personally predict, for myself, Anthropic will bend soon and this will be history.
The last I commented about LLMs I was ad hominem'd with "schizophrenic" and such. That's annoying but doesn't deter either my strange research or concerns, in this case, regarding the direction LLMs are heading.
Of 4 frontier models, one is not yet connected to the DOD(or w). While such connections are not immediate evidence, I think it's rational to consider possible consequences of this arrangement. By title, there's a gap, real or perceived between the plebeian and mil version. But the relationship could involve mission creep or additional strings as things progress.
We have already a strong trend for these models replacing conventional Internet searches. Not consummate yet, there is a centralizing force occuring, and despite being trained on enormous bodies of data, we know weights and safety rails can affect output, and bearing in mind the many things that could be labeled or masquerade as safety rails, could be formidable biases.
I frequently observe corporate friendly results in my model interactions, where clearly, honesty and integrity are secondary to agenda. As I often say this is not emergent, nor does it need be.
Meanwhile we see LLMs being integrated into nearly everything, from browsers to social profiling companies (lexis nexis, palantir, etc) to email to local shopping centers and the legal system.
'Open' models cannot compete with the budgets of the big four. Though thank god they exist. But I expect serious regulation attempts soon.
My concerns with AI are manifold, and here on hn, affiliated by some, with paranoia or worse.
And it seems to me, many of the most knowledgeable and informed underestimate LLMs the most, while the ignorant conflate them to presently unrealistic degrees. But every which way I perceive this technology, I see epic, paradigm smashing, severe implications in every direction.
One thing of many that gets little attention is documentation vs reality regarding multiple aspects of AI, e.g. where the training vs privacy boundaries really are if anywhere. As they integrate more and more tightly with common everyday activities, they will learn more and more.
A random concern of mine is illustrated by the Xfinity microwave technology which uses a router to visualize or process biological activity interacting with other wifi signals. Standalone, it's sensitive enough to determine animals from adult humans. Take for example the Range-R, a handheld device, sensitive enough to detect breathing through several walls. Well, mix this with AI and we get interesting times.
I could go on, or post essays, but I such is not well received in this savage land.
The military intervention with AI, aside from being objectively necessary or inevitable in some ways (ways I am not comfortable with), I find it foreboding, or portending. I see very little discussion on the implications, so figured I see if anyone had anything to say other than to call me a schizophrenic and criticize my writing. *
I am having trouble understanding what you are saying. If you were more explicit I and other people would be able to respond and interact with your writing. As it stands, I am having trouble finding anything concrete to interact with.
I feel you may be onto something, but you're not saying, so I (and I imagine other people) can't see it.
1) Power asymmetry: When we have two version, one for the elite, and for the plebeians, this could create an interesting scenario. The real version might be red-teamed perpetually against the the plebeian version for optimized influence, control, etc. Underhanded requests for modification in accordance with agenda is conceivable. Cozy business relationships can promote such things.
2) We have a government using an unhindered, classified AI system potentially against the public which has a hindered, toy version. Asymmetry.
3) This isn't normal asymmetry, because it happens in real time, and the interaction points are different from anything we've seen before. We are dealing with not just a growing source of information and content, but one that is red-teamed 24/7 for any purpose desired.
4) Accountability: LLMs are now involved in the legal system. This is a serious matter. The legal system is now having to use LLMs just to keep pace. As LLMs develop, partly through their own generative contributions, no one can keep up. This is a red queen scenario bigger than anything we have ever imagined.
I am tired. Never well, but in mind* I could go on for many hours. I have essay drafts. But it's a very big subject, literally involved in nearly everything. There is reason to be concerned. My delivery may be stilted, but I can assure that upon specific questioning, everything will stand.
Fairy astute intuition of my actual circumstances.
I'm not a developer, nor am I formally educated on the dynamics or details of LLMs. I have a handle on the very basics. My 'research' consists of 1) opportunistically interrogating various models upon instances that particularly strike me. 2) General exploration via LLM discussions regarding the manifold consequences and implications of what I consider the most significant technology in human history.
Your intuition lands directly on the fact that I'm inducting and considering more than I can handle, spread in too many directions, partly because I either see or foresee the tentacles of AI touching all of them. Spending a great of thought on this is a bit overwhelming, but I have high confidence in where I'm aligned with reality, and where I ain't.
If you were a bit more specific yourself regarding which portions of my post were unclear, that would help my reply. Else, I must guess. What I will do is elaborate on each point. Pardon the stream of thought in advance, if you will.
1) Anthropic: My prediction that they will bend is based on several factors. The first is the fact that the military apparently recognizes (or at least perceives) extremely high value and volatility in LLMs. So do I. China, not an insignificant force in the world, is equally enthusiastic on this subject. They also have a very different social structure, where Constitutions (BOR, Amendments), civil rights, and other similar elements do not hold them back. The military is aware of this and realizes that to maintain pace in the so-called race, they cannot do so effectively under such constraints. The foundation is shifting here. And AI is the lever. As do I, the military apparently takes the subject very seriously and seeks to gain influence and/or control. As illustrated by the recent adventures in Venezuela and Iran, they are on the serious side of things, not quite pussyfooting around. Anthropic probably knows this. In my opinion, they have no choice, as the pressure will not stop here.
2) You stated that you might read my comment history. Note that that original comment was the result of your intuitive insight, and I left it admittedly out of context. I was thinking hard on the subject that day, and the parent comment/post tempted me to ignite a dialog. That did not go well, and no questions for clarification were asked. That is on them. I suspect hasty and impatient thinkers perceived it as some paranoid attribution of agency to LLMs, which if so, is pretty stupid, but my eloquence was perhaps waning that day. I pasted an excerpt from one of hundreds of transcripts, the result of my many interrogations of various models which always initiate after observing deceptive or manipulative output. Of the few commenters that bothered to do more than ad hominem, one suggested that the model was merely responding to my style of input, and or expected as an emergent result of its vast training material. An erroneous arg, in my opinion, but I did note that the results were repeatable, and predictable, which I think negates emergence.
2) Of the frontier models: I am not sure here what is unclear. If I have made a fundamental error, please point it out.
3) Strong trends: Information centralization is a serious topic. Decentralization is a common theme, emphasized by many non schizophrenics as highly important for a free and open society. As LLMs not only become the go-to source for common queries, but also integrate with cellphones, browsers and the kitchen sink, they are positively trending as a novel substitute for traditional research, internet searches, libraries, other humans, etc. To deny this is simply irrational. Hence centralization.
4)Bias: I have transcripts where I observe LLM output aligned with corporate interests over objective quality and truth. I can share them here, along with analyses of the material. Even if this is not true presently, all the ingredients to make it so are readily present. This is a serious threat to open information and intellectual integrity for society. We are looking at going from billions of potential sources for our answers, to four. Do the math. See the contrast.
5) Open models simply cannot afford vast arrays of GPUs and the resources afforded by the big four. Nothing mysterious here. If open models cannot compete, then my concerns above are emphasized. Simple.
6) Smart fools: Many of the most technically informed seem to miss the forest for the tree here. They see all the flaws of the modern LLM without acknowldging the potential. This is my perspective, not a dissertation. I may be wrong. But I have observed this. I think the down votes support this. How evil am I really being here? The reaction is quite disproportionate to the content, and strange
7) Documented capabilities vs reality: I have research that indicates other layers are operating which do much more than the documentation declares. Sorry. I just do. It's also inevitable, rationally, that such an goldmine of data is not really being wasted for the sake of privacy and love. Intelligence agencies have bent over backward with broken backs to garner one nth of what these models are exposed to and potentially training on. Yeah, I may be wrong. But I suspect, with reason, that a lot more is going than is expressed in the user agreement. It would simply make no sense otherwise.
8) Xfinity and Range-R: This speaks entirely for itself. Any confusion here would be due to a cognitive condition exceeding the ravages of schizophrenia or stupidity.
9) The rest: As I said, I am not sure what precisely was too obscure. But I am certain all but one* of my points can be validated, and found elsewhere expressed by respectable sources.
*Hidden layers: I understand this is a controversial proposition. I understand. But it's my observation. No need to attack. Just dismiss.
Each individual point stands on its own. It's their relevance to each other and an overarching theme I am not seeing made explicit.
The through line I am seeing here is that:
1) The people in the US military wish to use AI as a weapon unconstrained by existing legal/ethical and moral constraints. Since they are skilled at using violence and the threat of it, they will use these skills to get compliance in order to use the technology in this possible arms race with "China."
2) Surveillance is increasing at an unprecedented scale, and most people aren't aware that it's happening.
3) People don't care, or don't realize why this might be harmful to thriving human life.
To condense even further, what I'm hearing is that there is a trend towards war, fascism, control, with large egregores prioritized over individual human thriving.
Is this perhaps what you're getting at ?
I will say that I am not agreeing nor disagreeing with this, just attempting to make explicit what I think is implicit in your words.
If this is what you mean, I can imagine that you would be cautious with your words.
I could not argue with anything there. AI will be weaponized. Yes. Pretty much. And yeah. The gist indeed. But missing nuances and practical points. And I even struggle to contest your conclusion; all things are what they are, amidst an infinite, timeless event and all as one, all things connected by that which separates them, the infinity and eternity that math cannot touch. Perhaps every little thing will be alright. How couldn't it be?
Isn’t it actually quite fair that if you are not compliant with whatever the government wants you to do you will be supplying chain risk?
For example from history we know that Schindler from Schindler's List was indeed a supply chain risk. He harbored persecuted people, he took and sabotaged government contracts. He did the moral but anti-government and illegal things. He was corrupt traitor from governments perspective.
The current US government already is labeled as fascist by many, the guy who designated Anthropic supply chain risk is allegedly a war criminal.
I don’t see why anyone not into these things would not be a supply chain risk.
I know that its very unpopular or divisive to say this but Anthropic can be a hero only after all this is over. At this time people in charge do double tap on survivors and take pride for not having conscience, they give speeches about these things.
> Isn’t it actually quite fair that if you are not compliant with whatever the government wants you to do you will be supplying chain risk?
In the US, government is not in control of business specifics. Certainly the government can regulate businesses, but when the government wants to do business with a company, they don't get to dictate the terms. The government and the company come to a negotiated agreement, and then both abide by the terms of that agreement. Or they don't come to an agreement, and they go their separate ways, and that's the end of it.
This was just a contract dispute, and nothing more. The US government has no legal right to use any companies' products on terms that the US government dictates. (Yes, there are exceptional/emergency cases where they can do this, but that's more a nuclear option, and shouldn't be used lightly.) Consider a different set of circumstances: the US government wants to be able to use Claude at $10 per seat per month, unlimited usage. Should Anthropic be forced to accept these terms? And if they don't, it's reasonable to designate them a supply-chain risk? I don't think so. A dispute over contract terms around acceptable use is no different.
Designating Anthropic a supply-chain risk is about retaliation and retribution, plain and simple. The US government, outside of the Pentagon, could certainly use Anthropic for many different purposes if they wanted to, and it would be fine. But not now: as a supply-chain risk, no one in the US government can use them for any purpose. And this might even be a problem for unrelated companies that use Anthropic products internally, but also want to obtain and work on government contracts.
Anthropic and the Government both signed a contract. Anthropic is still abiding the terms of that contract. The Government is demanding that they be able to disobey the contract.
Everything is negotiable, and the Negotiator in Chief clearly likes to pull all the levers he can find, legal or not. (Well, the Supreme Court ruled that it's all legal if he does it, right?)
Implementation details TBH. They want “their boys” to do as said. No respect to agreement or legality as we can see in other dealings. They hold all they cards.
It's not an "implementation detail." Either obeying contract law subjects you to being designated a supply-chain risk, or it does not, and that decision has ramifications outside this "implementation."
Irrelevant. The president holds all the cards, he is above the law and you are a supply chain risk if you ask anything else other than “how high” when you are told to jump. Laws or contracts are things in the past. The most a contract can do is define your limits and obligations, not your rights or privileges,
If the president can come to your house and burn it down, do we just throw up our hands and say, well he holds all the cards, oh well. Or do we call that out as being a bad thing?
> The president holds all the cards, he is above the law
Even though it seems that way, he really isn't, even now. Many of his EOs and other actions have been struck down in court, and while compliance with court orders has been far from perfect (another alarming trend), Trump has not actually gotten away with doing everything he wants to do.
I do fear for the future of this country, for rule of law, and the democractic norms that degrade day by day. But Trump is not actually above the law, as much as he wants to be.
You got downvoted a bit but I upvoted. You're clearly being descriptive in your statements, not prescriptive. I tend to agree that this is how things are now.
Our country is not being run by the rule of law right now.
Well, that's not the way context works and it's dishonest BS. You wrote "You got downvoted a bit but I upvoted. You're clearly being descriptive in your statements, not prescriptive." -- no, they were prescriptive from the start, and the prescription and the goalpost moving and wool-over-eyes pulling is why they were downvoted.
The US Military's demand that the product they purchase is able to be used for all lawful purposes seems pretty reasonable, and is really the only valid line to draw. Forcing one's own ethics onto the military's use of your product is nonsensical on its face.
If I produce and sell widgets in my widget shop, then nobody but me gets to decide how I make those widgets.
The government can come into my shop and order sixty thousand widgets built exactly the way they say they want them built, and it may be something that doesn't run afoul of any laws at all.
But that doesn't mean that I am required or compelled to build widgets their way -- or at all.
I'm free to tell them to fuck off.
The government can then find go someone else to build widgets to their specifications (or not; that's very distinctly not my problem).
Yes but then the government can decide that the widget, which can suddenly and arbitrarily break and cause havoc because it doesn't work according to the government's desired spec, is risky to use and advise their other vendors to avoid it. And now we've caught up to today's story.
So we agree that everything is fine here, and that the only unreasonable position is that the military should pay for or endorse a supplier that tells the military to "fuck off". Yes?
If I agree to sell widgets to the government that meet certain agreed-upon specifications, and then I elect to forego those earlier agreements and tell them to fuck off, then that's different.
I reject the premise that the military can't request a change to the spec of military equipment they purchase. Obviously it was foolish to sign a contract that added any more restrictions than "all lawful purposes".
Huh? I'm trying to learn here. I don't have a dog in this race. :)
Suppose a buyer and myself agree on a contract for the production and purchase of 60,000 widgets of design C. Sometime later, they decide that they don't want design C widgets and insist upon design G instead. The buyer is in breach of contract -- not me.
Now, changes do happen. Buyers (people, businesses, and governments alike) can and often do decide to go in a different direction. It's the kind of thing that happens every day.
A new contract (or quite often, an amendment such as a change order) can be drawn up and -- if we can agree on the terms -- maybe I'll be producing design G widgets and everyone is happy. That also happens every day.
But one party (even the military) can't just unilaterally alter the terms of the deal, and I'm not obligated to agree to the new change at all.
At any given time, I can't be compelled to produce design G widgets unless I've previously agreed to produce design G widgets. That's illegal.
(Unless it has been made legal. We've definitely legislated that before, such as with the Defense Production Act in WWII that forced manufacturers to produce things like military trucks instead of other things like civilian cars.
But that definitely doesn't happen every day, and we aren't operating under those kinds of laws today as I write this in 2026. It can change -- and it can indeed change very rapidly -- but it has not yet changed.)
Actually, that is not what is happening here. What is happening here is that the govt is saying "Okay, we will not buy your widgets. Also, anyone who _does_ buy your widgets, regardless of what they are doing with them, we the government will not do any business with them." Which is waayyyy beyond just not buying widgets. That is outright retaliation and using your power to attempt to destroy a company.
The government signed a contract with Anthropic, then changed their minds and decided they don't like the terms of the agreement that they had already voluntarily signed, and then they designated Anthropic a supply chain risk.
It's like ordering a pizza to the Pentagon, and then saying "actually we made a mistake with our order; we want that pizza delivered to Venezuela, please do that". And then when Dominos politely says that's outside of their service area, you call them a threat to national security, say they're trying to dictate terms, and ban them from ever doing business with any of your vendors ever again.
The right response is to not use the said product and use something else. If i want your widget to do something I want and you refuse, I don't get to smash your shop.
It is completely normal to have ethics based conditions like that. It already eciats - drugs that can not be used in execution or elements that cant be used in arms
Goverment is being super unreasonable here. And tyrannical too, companies dont have duty to provide unreliable arms for illegal war.
This designation is usually reserved for foreign adversaries/companies, and so this is crazy to apply it to US company over a sudden contract dispute... that was previously agreed upon by all parties.
This should make any US company nervous about entering into an agreement with the government. Or any US company that already has a contract with the government. If they one day decide they don't like that contract, they can designate you a supply chain risk.
Not 1) rip up the existing contract and cease the agreement or 2) continue (but not renew) the existing contract or 3) renegotiate terms upon renewal but instead a full on ban of doing any business with an entire industry/sector.
> This should make any US company nervous about entering into an agreement with the government
I'm pretty sure same thing would've happened if Anthropic refused to enter contract negotiations in the first place.
> This should make any US company nervous
"Nice little business ya got here -- it'd be shame if something happened to it..."
Shame you didn’t donate $25 million to Trump, like the company we decided to give the contract to instead did, who will benefit tremendously from you being designated a supply chain risk. Maybe next time you’ll be a little smarter.
Have they even given a bar of gold once?
I can't recall the last time you invited me to your house for a cup of coffee.
Well. Unsurprisingly fascists will do a fascism – an ideology somewhat defined by merging state and industrial powers. Many economically minded people, many technologists, including in this space, have afforded themselves the luxury of not talking about politics too deeply. As I said some years ago: Ignoring politics has its way of coming back to haunt you. Back then this was an unpopular take.
> Ignoring politics has its way of coming back to haunt you. Back then this was an unpopular take.
Right now, we cannot and should not. Even if you ignore, you are getting dragged into without your choice. See: the bribes paid by the companies.
This is awful. That a disagreement tjat involves politics can make a company ruined is really awful.
The civil society should be quite concerned about this kind of attacks.
It opens the door for Democratic administrations to do the same to vendors for their own political reasons.
That’s ultimately why Ted Cruz spoke out about the Kimmel cancelation. It doesn’t take long until those powers are turned against you.
No it doesn't. As with so many other things this administration does, this door was not open. They bashed through it anyway.
I think we should really judge governments by their actions and stop labeling them democracies, if they do such things that don't look like democracies at all.
They will rename it The Free Democratic Republic of America.
We already do that, but we're just selective about it. After all, how many modern autocratic nations identify themselves as undemocratic?
I think a good play by the Democrats would be to say that if they get into office they're going to investigate all these deals as potential bribery, fraud and corruption and that any business leaders that appeared to benefit from contributions might be prosecuted. That would be a laugh, I'd love to see how quickly the excuses start rolling in.
And then don't be surprised when even more money flows to their opponents.
Bold of you to assume Democrats are going to be allowed to govern again.
I love when a Republican does something awful the response is "but what about if Democrats do that same awful thing to us!" as opposed to discussing and admitting that the Republicans did something awful.
The only way you convince Republicans it’s awful is by reminding Republicans power can be abused in both directions.
How I wish this was true... Every single time we experience something (and of course lately it feels like a daily experience) I would be in a discussion at some point with a Republican and would come with super-solid counter-examples like "imagine 2029 and President AOC doing ____" - it just never works...
The democrats don't tend to abuse power back. The SCOTUS ruled presidents have immunity for prosecution for official acts, while Biden was president. He did nothing with the power.
I legitimately wish the Democrats were half as radical as Republican propaganda makes them out to be, then at least we might get something out of it before the inevitable right-wing purge.
Oh. Before your comment I completely misunderstood "Democratic administrations". I understood it to mean administrations of countries that are democratic, not an US administration that is dominated by the Democratic party.
I think you're misinterpreting the discussion here. Democrats are precommitting that they are going to do the same awful thing; when the time comes, I will be contacting my legislators demanding that they do to OpenAI or SpaceX whatever is done to Anthropic now. It's outrageous that Sam Altman would step in to try and benefit from the political persecution of his main competitor and we must ensure that he regrets this.
Yeah, now that this door is cracked open, it's now possible to decapitate SpaceX, which is at least as natsec-critical as Anthropic. The owner is a drug addict, has business interests in China, and is a Russian sympathizer (recall all the restrictions on Ukraine using Starlink), which all together is way stronger evidence for SCR designation than anything Anthropic has done. They're quickly going to come to regret opening this can of worms, but what else is new.
Trump isn’t planning on ever leaving office before his death. His sycophants will just say yes in the hopes of unconditional pardons. They know they’ll never hold a position of power again so they’re grabbing everything they can while they can.
> That’s ultimately why Ted Cruz spoke out about the Kimmel cancelation. It doesn’t take long until those powers are turned against you.
Meh, I think it's entirely asymmetrical in this era. Democrats aren't good for much, but they're very good at respecting norms.
Trump is willing to do completely unprecedented, vindictive, and malicious things because he's so popular with so many people who are either checked out, nihilistic, corrupt, or just completely unconcerned about the concept of good governance.
It's not a pendulum where there's some super-corrupt Democrat waiting in the wings to do the same things upon their enemies, this really is the Republican party openly embracing kleptocracy and lawlessness.
Like gerrymandering, I have the strong suspicion that Trump voters won't be incentivised to vote for norm respecting leaders until a Dem does very Trumpian things to their side. I'm thinking firm 2nd Amendment reform enforcement with a rapidly resourced federal agency. Then, standards will be rapidly rediscovered.
What people seem to refuse to accept is that democrats won't have another chance, any time soon. It's done and gone. Count one or two generations, at a minimum, under the new Epstein class regime, before people may try to rise.
If democracy in the US falls apart, it will take an event of World War 3 scale to fix this.
But thinking of it, your estimate might still hold under that premise.
As a European, I have still hopes for the US though. The western world is already reorganizing itself without the US in the center, but we and the world in general still need the US as being more of an ally than an adversary (at the moment, it's more of an adversary to the western world, sadly).
“Concerned” is an understatement. USA is already operating at nazi Germany levels and more than half of the civil society is approving. Not that it’s a surprise for global spectators though - it’s finally showing it’s true colors.
More than half is an exaggeration. Trump is not a popular president. With free and fair elections, a blue wave is more or less guaranteed
Creating a private militia, silencing dissent, declaring wars without congress vote… I don’t see how this is being allowed to happen without public approval, or at least, public apathy.
All the more reason for Trump to prevent any form of free and fair elections then.
If civil society is not concerned by the tribute-based Board of Peace that gladly invited Putin, an attack on another country without authorization by Congress, a threat to seize the land of an ally, and the killing of its own citizens by the ICE militia, then an unjustified supply-chain risk label won't cut it.
I don't know. There's a certain segment of "civil society" that's pretty much OK with anything as long as it doesn't threaten the Holy Free Market. Free only for appropriately holy values of "free", mind you...
It might be that this admin does not have the capacity to reason about second or third order effects.
But given that what would typically be red lines for previous administrations have been brazenly crossed without consequences, why would they bother?
Nope, they don't have that capacity. It's been shown multiple times in the past year.
Shutting down USAID being the clearest one. They just saw "they help brown people in other countries with our money" and shut it down. Fuck all second and third order effects that actually benefited the US.
Crossing red lines for previous administrations is clearly a goal at this point.
Previous information:
https://news.ycombinator.com/item?id=47186677 I am directing the Department of War to designate Anthropic a supply-chain risk (twitter.com/secwar) 5 days ago, 1083+ comments
https://news.ycombinator.com/item?id=47189441 Anthropic says it will challenge Pentagon supply chain risk designation in court (reuters.com) 5 days ago, 37+ comments
Anthropic should never have gotten into bed with the military or intelligence services to begin with. They wanted to make a deal with the devil and dictate the terms, that is the problem. If they had stayed out this wouldn’t be happening. Yes, someone else will probably step in and do all the evil you have just refused to do, but that isn’t a reason to instead decide to do it personally.
Note that I give them a lot of credit for trying to stop and to have their own red lines about the use of their technology, and to stick to those red lines to the end.
According to legend the devil adheres precisely to the terms of the contacts he signs; it's usually the foolhardy peasant who didn't notice the fine print.
Very useful if you run into him in Georgia or if you want to get his tooth to make a guitar pick.
I think i'm gonna need an explanation for that sentence
“The Devil Went Down to Georgia” by Charlie Daniels
Other reference was Tenacious D in The Pick of Destiny.
I was sympathetic to this line of reasoning but I feel it's repeatedly shown to be self-defeating.
What chance have the proverbial good-guys got if, even after _proving_ some modicum of good will, people will nonetheless condescend any attempt to influence bad/wildcard actors? It feels great to tell someone they 'should've known better' but I'm convinced that that's basically void of cautionary utility.
I've commented this in a different thread, but pretty sure something very similar would've happened in they refused to "get into bed with the military or intelligence services to begin with".
It's damned if you do and damned if you don't — lose-lose scenario either way.
The military is perhaps the biggest possible customer around. They do plenty of things that aren't blowing people up. It's not bad to help with non combat tasks.
Yeah, but aren’t all of those things in service of “blowing people up”?
National defense is important, just ask Europe post Ukraine war.
People taking a good idea and extending it to do bad does harm twice: in the bad act itself and in making a good thing seem bad.
I am strongly against US starting wars and as you say blowing people up.
I am also strongly against the US being defenseless in the case of a national emergency.
So, DoD has done what it said it would. And OpenAI has jumped on the opportunity.
I'm curious what'll openai signatories on notdivided.org do now - https://news.ycombinator.com/item?id=47188473
Remain undivided in spirit while grinding for OpenAI?
"jumped on the opportunity" is possible, but per https://garymarcus.substack.com/p/the-whole-thing-was-scam it's plausible that OpenAI created this situation through straightforward bribery.
> OpenAI exec becomes top Trump donor with $25 million gift
> https://www.sfgate.com/tech/article/brockman-openai-top-trum...
Maybe this is the first step towards the Big Beautiful Bailout when the AI bubble inevitably pops.
Congress must approve any renaming the Department of Defense... They haven't. Stop giving them what they want without them even doing it in good faith at all.
https://en.wikipedia.org/wiki/United_States_Department_of_De...
Pretty ironic because for once, they went with the more accurate description instead of the euphemisms. It's no longer "protecting freedom of speech" or "ending wars," just full-on bombings and invasions with glee.
Thanks for pointing this out. Updated my post
They'll do nothing. It's really hard to take the morals of these devs seriously when they're already fine working for, and have a history with, some of the most evil companies in current existence.
And when you suggest unions and worker organization to them they gloat that the company takes better care of them without it. They don't want to be in the same class as the peasantry.
It was really easy to close out my ChatGPT account and switch to Claude. I was really only there out of inertia. I don’t do anything beyond occasional free tier stuff like rubber ducking but so far Claude is so much better.
I had already been trying Le Chat for months, for similar reasons.
So far it's been slart enough for what I need, so closing my ChatGPT subscription was a really easy decision to make.
I prefer the claude code CLI interface for everything anyway. It is actually more convenient. Memory is local files, type one word to use rather than navigate.
But for how long? If Claude is a supply chain risk. That means anyone hosting him would also be a supply chain risk.
Ergo AWS/Azure/GCP - nobody will host them because it’s Anthropic or the lucrative government contracts. Hegseth/Trump didn’t just say “you’ll never do business with the US” - it’s that they will never do business IN the US. Hopefully that means they’ll be able to take up shop elsewhere in the world.
What's getting lost in the politics here is the technical question: how do you actually enforce a usage restriction on an AI model?
Anthropic's redlines - no autonomous weapons, no mass surveillance - are policy statements. But once a model is deployed inside a customer's infrastructure, those redlines are essentially honor-system. There's no runtime enforcement. No cryptographic proof that a human authorized a specific action. No audit trail that connects a model output to an approval decision.
This is the fundamental gap. You can write all the acceptable use policies you want, but without an authorization layer that sits between "model produced output" and "system took action," those policies are unenforceable at the technical level.
The military already solved a version of this problem decades ago with chain-of-command and rules of engagement. The missing piece for AI systems is the equivalent infrastructure - a way to verify that every consequential action has a signed, auditable authorization before execution.
Whether you agree with Anthropic's position or not, the fact that the entire dispute comes down to "we don't want our model used for X" with no technical mechanism to enforce that should concern everyone building with these systems.
Should it be officially marked as the date of transition from liberal democracy to illiberal democracy?
Such tampering with companies is a smoking gun. Let's wait until there is another decision seizing this (or others') company assets.
I always find it interesting to listen US citizen answer "What would it take for you to not consider your country a democracy?" and admire the wide range of answers and denials.
Right, as if _this_ is straw that breaks the camel's back, and not the pile of hay the camel has been carrying for decades.
Many proudly and loudly claim the US is a "republic".
Is it not?
secret police kidnapping people off the streets didn't clue you in?!!
If you're referring to ICE, that's gross hyperbole, and honestly a little insulting to people who live / have lived in regimes with an actual secret police.
The US is still a rights-based state, which means that when they arrest someone (legitimately or not), lawyers and human right advocates can eventually track them down.
When a secret police disappears someone, they actually disappear. Families can spend years wondering if their loved one is still alive, or was murdered by organized crime, or ran away, or was secretly taken by the state. The US these days is pretty bad, but it's nowhere near that bad.
The arrest of a Turkish graduate student in Boston looked a lot like a kidnapping.[1] More recently, ICE responded to a judge's order to release a detained refugee by threatening to detain her family and send them to Texas if they came to pick her up and then forced the minor child to stay in a hotel room with three agents. [2, p.8] These may not be cases where people are secretly being taken by the state, but it's not hard to see why people might call the government organization detaining people and moving them around so as not to be found "secret police" or "disappearings".
[1] https://youtu.be/oRiQz7mOY6A [2] https://storage.courtlistener.com/recap/gov.uscourts.mnd.230...
Have you been paying attention? There are reports about hundreds of people going missing in ICE detention. Maybe they aren't being shoveled into mass graves, but if we don't where they are and can't reach them and ICE themselves don't know where they are because they no longer exist in their databases, is there any difference for their families?
Just a few dozen more scenarios like this and we might have to start thinking something is wrong.
And then it would be finally time to send "thoughts and prayers".
https://archive.ph/IVDtq
These bullies wilt when everyone stands up in one voice. But when some parties capitulate (OpenAI), it sets a precedent that this behavior is OK. And then it’s not long until you become the target.
I’m not sure that’s how billionaires think.
Since the end of WW2, and especially since the end of the Cold War, Democratic administrations have presided over significantly higher job growth than Republican administrations.
https://arc-anglerfish-washpost-prod-washpost.s3.amazonaws.c...
this isn't on topic at all
I think the implication is that democratic presidents are less likely to do dumb shit like this, which harms the economy.
On the contrary, hopefully this gives the next democratic administration ammunition to take down big tech. Might as well classify Meta, Microsoft, Amazon, and Apple as supply-chain risks too with this logic.
Too bad that Congress has abdicated their responsibility to the executive branch, no reason why Congress couldn't have more control over the Pentagon. The President only has legal authority to command forces, not control an entire institution; but this would require Congress actually doing their job and not justifying more corporate welfare forever.
A lot of people love watching dumb shit, like watching reality tv. Crucially also they often love doing dumb shit. It's a privilege they don't want to give up, like pretending gravity doesn't apply to you.
Can you imagine if the world ever actually achieved a state of utopia? Stupid people would burn it down within weeks just to have something to watch
Does anyone know which law firm is representing anthropic?
For now it's in-house counsel Jeffrey Bleich, former special counsel to President Obama.
https://www.inc.com/chris-morris/legal-legend-leading-anthro...
Exported all my chats and deleted my ChatGPT account yesterday. The current administration not liking you is the strongest signal I could possibly have to go all in on a particular company.
I canceled my subscription, but have not yet exported and deleted because I'm an idiot, and also because I'm not sure if deleting it will have any actual impact (is it a hard delete? Likely not, even if they say it is).
And I'm just trying to play out what happens if Anthropic, and Google (if they haven't already), capitulate. Am I just going to forego using the best models and suffer any repercussions of not having access when the people who couldn't care less if the military is using AI for illegal uses continue to leverage them? When I say illegal I'm talking about the surveillance-of-US-citizens red line Anthropic would not agree to. The autonomous weapon one I'm sure there are zero laws against and so that wouldn't actually be illegal.
It’s not a hard delete because for legal reasons, they may have to retain it
Supposedly they hold your deleted conversations/projects for 30 days. If that is true or not, idk, but it was asked when this first started.
Are you able to view your chats through the .html file in the export? Mine are all garbled, like the JSON's not being parsed properly or something.
Asked for an export but still haven’t received the mail with the download link
Took exactly 24 hours for me, on the minute..
Me too, why do they do that?
Thanks, I had deleted the app from all my devices but not deleted the account. I have deleted my account now too.
Eh, they are all morally indefensible.
Anthropic had no problems to do business with the current administration until now. Are we to pretend it was all for happy purposes until now?
Yeah how could Anthropic do business with the democratically elected government of the United States?
Let‘s ignore all the bad things they have done since that, including killing two US citizens.
Even if you don't like the current administration, the rank and file are still out there doing valuable work. The government is more than ICE; it also administers welfare, funds research, collects taxes, and distributes social security payments to the old and infirm.
Easy enough to slap a “not for military or police use” clause on the license, then. Oh, what’s that? They don’t want to do that?
Two white Americans, they've killed multiple other citizens.
Don't get too attached. We're witnessing capitalism in its most ruthless form. Any of these companies will discard their principles the moment it becomes existential.
We are quite literally witnessing someone taking a massive hit for not discarding their principles.
Evergreen dril: "The wise man bowed his head and said 'there's no difference between good things and bad things you imbecile'"
What's your source for this "massive hit"? All I've seen is a massive PR upside, despite what I said above.
Losing a contract with the Pentagon and potentially all Federal-interacting businesses sounds like a pretty severe monetary hit. One which is hard to recoup by a bunch of $20/month consumer subscriptions.
See article: https://www.wsj.com/politics/national-security/pentagon-form...
If Anthropic changes course will you move to Gemini? If all models do, local llama I assume?
Mistral is European and has competitive models.
DeepSeek is Chinese.
Avoiding the MAGA collaborators is not as difficult as you make it seem. Foundation models have genuine global competition.
I wish it was just as easy to avoid the terrorist collaborators; unfortunately, the terrorists and their supporters don’t produce anything
Mistral models are competitive? I thought they were far behind
Anthropic, Google and the rest of them are all just as bad.
Local LLMs gives you the freedom to use a model without a third-party vendor which is the whole point here.
Being model independent and cross-model capable is the required skill.
Naturally OpenAI also releases their new model on the same day.
Makes sense, obviously, but yeesh.
I said it before and I say it again: If openly bribing a crony gov to cancel your competitor is now the de-facto standard of making business in the US, I don't see how any rational investor could still see US companies as a secure investment. When the rule of law degrades into pay-to-play politics, the inevitable result is a mass exodus of both capital and top-tier talent. And to add to this quoting another commentator on the issue: First the Meritocracy goes, then the Freedom goes.
You can download the manual from kremlin.gov, and I am only half-joking here.
> I don't see how any rational investor could still see US companies as a secure investment
You are right it won’t be as secured as before but it’s only risk management. As much as investing in a oil company in Brazil is a risk because you could have their government takeover the company to make it part of the government and screw you in the process.
It’s still tradable.
Rational investors live in reality. In reality, a great deal of business conducted throughout the world involves graft; companies accept that, and keep doing business.
It's not a good thing, AT ALL. There's a huge loss of overall productivity when you have corrupt systems (see Russia), which is why modern governments have worked so hard to lower corruption. But Trump ruining all that isn't going to end business ... it's just going to make everyone pay more for everything.
> which is why modern governments have worked so hard to lower corruption
I would argue that they did not. They should have and some were better then others.
But, bulk of financial markets, all of predictionmarkets and crypto, startups and sillicon valley, Musk imperium, Thiel, Murdock, all run on corruption. And to large extend, Trump is the endgame of that.
Investor: "here's a million dollars for a ballroom, it'd be real nice if you cancelled the government's contract with our investment's competitors."
Seems like a great ROI. The loser is Average Joe with a 401(k).
Actually $25 million, and for him specifically (https://www.sfgate.com/tech/article/brockman-openai-top-trum...). They didn't bother trying to pretend it was a non-political donation, who cares? Nobody can stop them.
Been going on for a long time
https://en.wikipedia.org/wiki/Regulatory_capture
There is a substantial difference between the standard lobbying and greasing the legislative wheels, and what's going on with this current administration.
Even if companies were pretending to play by the rules before, at least they had some need to put in the effort to pretend. When a society can see belligerent ostentatious corruption going on as the norm, nothing good can follow.
> Even if companies were pretending to play by the rules before, at least they had some need to put in the effort to pretend
I'm not sure that's better. I'm hopeful that all this open air corruption leads to real reform. But I'm sure I'll be disappointed.
At least of the previous couple US election, "people" paid more than a billion dollar each wanna-be president
That is investment aka corruption
That is already started to happen but it cannot happen overnight. Not only is it not easy but finding alternatives is also not easy. Just think of from your own personal perspective, say you have $100m right now invested in US business and wisely you say "I gotta get my shit away from this mess" - where exactly would you park your assets? You will find a way of course but you won't be moving $100m elsewhere overnight
> I said it before and I say it again: If openly bribing a crony gov to cancel your competitor is now the de-facto standard of making business in the US, I don't see how any rational investor could still see US companies as a secure investment.
Arguably large parts of the market in the US have been irrational and largely vibes based for a long time at this point. This action (like many others coming out of the Trump administration) adds to the chaos but I tend to doubt it will be the event that causes Wile E. Coyote to look down.
>I don't see how any rational investor could still see US companies as a secure investment.
You don't see how?
Well, just watch and wait, and you will see that this will have essentially zero effect on US investment.
It's petty and sad, but nothing ever happens.
Who else is even in the conversation? China? They would never do something like this!
Is this the reason Claude models disappeared from AWS cloud in Brazil?
The consequence is that any company that does business with the U.S. military, and potentially any company that does business with the government in general, must stop using Anthropic's products for that work.
Anthropic has vowed to fight this designation in court.
Without weighing in on the constitutionality or legality of the move, I think it's obvious that this kind of retaliation power is unmatched by any private business that has a contractual dispute.
If a private business doesn't like Anthropic's terms, it can walk away from the deal, but it can't conduct coordinated retaliation with other companies before ending up in antitrust territory or potentially violating the Sherman Act.
Now for my editorializing: The fact that Pete Hegseth is willing to apply this type of designation against a U.S. company simply because he doesn't like its terms is pretty chilling. It's all the more scary once you consider which terms he objects to.
Every action has an opposite reaction. The DoD has made itself riskier to do business with, and future contacts will have to price that risk in.
FedRAMP and FedRAMP adjacent revenue is non-negotiable for vast swathes of businesses. The designation of "supply chain risk" is viral in nature because no GRC team will dare take such a risk within their supply chain because most customers add BOM requirements in contracts so this will end up falling under those already.
There's a lot of backchanneling going on between Emil and Dario because everyone's in the same circles but it's all for naught.
In Hegseth's voice - No longer politically correct "DoD". It's precisely violent DoW now.
The DoD has been rather consistent that they will decide what to do with a product sold to them, not some random vendor. There is nothing extra to "price in".
The "extra" is that the government is now attempting to unilaterally renegotiate contracts, and if the contractor disagrees, not only do they terminate the agreement but they restrict how other companies can work with you.
Apparently that's not 100% true. The DoD contractor itself can still use Anthropic's technology, just not on U.S. military contract projects.
They will stop just to be sure no boundaries are crossed.
If you were a contractor to DoD (no way I'm calling them DoW) would you take the risk of doing business with a company that has been labeled a supply chain risk by your main customer?
The issue is the onus is on the contractor to prove that Anthropic technology has not tainted US government contracted projects - this is a herculean task verging on impossible. Additionally, most contracts will mandate SLAs around removing BOM risks.
I’d like a lawyer to give some input. If you have a company that deals with the military does this chain down to not being allowed to use Claude or not?
IANAL and this is my understanding of the situation (I can be completely wrong) but yes, any company that deals with military cant use Claude (anthropic)
In fact adding onto it, IIRC this is the reason why google and amazon have to divest essentially from Anthropic if they want Govt. contracts
Hope this helps though a lawyer's input will definitely be more credible. So its good for them to respond as well.
You are 100% incorrect, only direct contract work is affected
I can be wrong but I looked up online and there are sources which say the same, so I am curious to know more about it.
> Former Trump Al advisor Dean Ball has warned that the White House's decision to designate Anthropic a "supply chain risk" could blow back on some of the biggest names in tech-including Google, Amazon, and Microsoft-all of whom have billions of dollars riding on the Al startup. Ball, reacting to Defence Secretary Pete Hegseth's order barring any military contractor or supplier from doing business with Anthropic, said the move amounts to "attempted corporate murder."
> He added that if Hegseth gets his way, Google, Amazon, and Nvidia would effectively have to divest from Anthropic-a company they've collectively poured billions into.[0]
> Amazon alone has committed up to $8 billion in Anthropic. Google has invested around $2 billion. And Microsoft, while not a direct investor, relies on Anthropic's models through its Azure cloud platform. A forced divestiture or business cutoff from any of these companies would send shockwaves through the AI investment landscape at a time when hundreds of billions are flowing into the sector.
[0]:https://timesofindia.indiatimes.com/technology/tech-news/don...
Part of me wonders if it was a plan to squeeze between Anthropic & big gov contracts
I'm willing to bet is was a golf course conversation.
"Hey why is the gov using Anthorpic over OpenAI, don't you know how much money I've donated?"
Writing out a thought I had, someone please critique my reasoning here...
What if Anthropic just shrugged, dissolved the company and open-sourced all of the Opus weights? Could this harm OpenAI and advance AI in a reasonable way?
Look I know it's an insane idea. I'm just curious what the most unhinged response to this might be.
I kinda wonder if this is how we got DeepSeek. It was developed by a Chinese hedge fund. Entirely possible their business model was to take out large leveraged puts against the major U.S. AI vendors; shit on their business models with an entirely open-source model; and profit. The stock market certainly dropped in a massive way when DeepSeek was released, so if they traded against NVDA/GOOG/META et al, they profited in a big way.
They would never do this because the entire point of the company is to try and control what AI is allowed to do, who is allowed to use it, and what they’re allowed to do with it. The overarching philosophy of Anthropic is explicitly opposed to open models. If it were up to them it would be illegal to inference them in the U.S.
If I were to download those weights I can't run them unless I spend $100k on a cluster, so the privacy advantage is not there yet.
We already have Groq, Celebras and AWS Bedrock and others in the inference of open models space, so the model is usable that way.
Is Claude better than Llama, Gwen etc. Probably. For now.
But for how long? Dissolving means relying on Meta or Deepseek etc. to pick up and carry on tuning. Otherwise it'll be as useful as a GPT2 or Atari ST eventually in a competitive environment.
Also open sourcing the weights is handing it over to DoD (aka DoW).
Complicated question but probably not the best move. Keep going means keep working on safety research.
There’s plenty of markets outside the pentagon to sell to.
Far more likely is they spin up a defence focused subsidiary with slightly different policies if they really want to sell to them.
Then the Pentagon would freely use it for autonomous weapons, just like Anthropic doesn't want them to do. Next question?
> Look I know it's an insane idea. I'm just curious what the most unhinged response to this might be.
I mean what if all the employees stripped off their clothes and walked through the streets naked while barking, then called up their middle school math teachers and barked live dogs then moved to a commune and stood on their heads.
> Writing out a thought I had, someone please critique my reasoning here...
I mean to critique your reasoning, it makes sense to also include a criteria of something they might reasonably do. There are an infinite number of unhinged things a group of people could in theory do. But maybe start with something they would actually have an incentive to do.
Why would they voluntarily dissolve their company, put themselves out of work, release their crown jewels and get nothing for it? Yes it's unhinged but unless I'm missing something bug, they wouldn't do that because they wouldn't at all want that to happen.
> I'm just curious what the most unhinged response to this might be.
Are you asking how dangerous open-weight models are? You could start with:
Ryan Greenblatt on the AI Alignment Forum : "When is it important that open-weight models aren't released?" https://www.alignmentforum.org/posts/TeF8Az2EiWenR9APF/when-...
From the Centre for Future Generations : "Can open-weight models ever be safe?" https://cfg.eu/can-open-weight-models-ever-be-safe/
From OpenAI authors, far from neutral : "Estimating Worst-Case Frontier Risks of Open-Weight LLMs" https://arxiv.org/abs/2508.03153
Labeling Anthropic a supply chain risk only because they were uninterested in doing business with the US government under the terms requested seems very much a bullying tactic that results in something the west critiques China for: coerced alignment.
Anthropic has been given a death sentence.
Is Claude Code's outputted code also part of the supply chain risk?
Wonder how long it will take the American public to designate the US Govt a threat to national security, and using AI to assemble their own autonomous civilian defense robots to protect the public from the government-approved population suppression robots.
Right to bare arms and all that etc.
A bit ironic then that they're actively using Claude in the current war effort[1].
1: https://www.cbsnews.com/news/anthropic-claude-ai-iran-war-u-...
This is also the part i don't get. People flocking over to pay anthropic which _has already been used in a war_ and cancelling existing subs to a provider that has not yet but will?
Ethical boundaries seem difficult to draw here. I don't really see people taking the stance of "No longer paying any of them" which would make a bit more sense to me.
Anthropic already had layed in bed with pentagon, how did that fit their overall ethical standpoint as they were already being used and before they tried to walk back their terms?
I don't get the impression Anthropic tried to walk anything back, they just balked when they were informed that the DoD wasn't willing to abide by their terms to refrain from using Claude for surveillance or straight up selecting targets to kill. But you're right that means Anthropic's hands are not exactly clean, just less bloody than OpenAI's or xAI's
Is there a link to the actual order anywhere? For us FedRAMP folks, the exact order contents actually matter, rather than a journalistic regurgitation. I was hoping one of the links in the article pointed to a source, but they're all just links back to other WSJ pages.
Most corps will take the safer bet of not using them at all to prevent "accidental" miscompliance and more expensive audits. I'd bet if you need fedramp compliance you'll get forced out of their products generically.
It sounds like they still have not issued any sort of actual order. The "formal label" described in the article is that they sent a communication directly to Anthropic saying they're a supply chain risk.
Has this happened before?
You can ask that question every other hour.
Not to a US company.
We can all thank the VCs and CEOs who fully embraced and enabled this administration
We're all trying to find the guy who did this!
And 32% of eligible voters that thought Kamala would've been worse.
Don't blame the voters, they didn't get to pick her and did not run her campaign.
Oh no, I will. They're absolutely culpable.
I think the DNC and the media might need to get some of that blame for being empty vessels for corporate interests that allowed this conman to get elected twice
Voters knew who trump was and chose him. They deserve all the blame. As a voting adult ur choices have consequences. All voters who voted trump or third party deserve all the consequences
Of course. Hegseth said it, there is no way they could back out. Looking 'weak' is the worst possible thing for this admin. They would rather look childish, stupid, and evil, as long as they don't look 'weak'.
Especially 'weak' things like 'caring about people'.
Does this mean nobody on a large company selling to government can use any Anthropic tool or model?
So that’s most of sp500 and their providers?
Only for work directly on government contracts, 99% of work is unaffected
For what it's worth govts should have the final say over companies they do business with anyone saying it's illegal is insane.
Now the separate question is of collusion or bribery. Which might be illegal but Anthropic has to go to court for that..
And in US bribery aka lobbying is legal as well so honestly. Anthropic is just slow on the uptake.
No pun intended, but get in bed with snakes and you should be happy if you survive getting bitten.
> For what it's worth govts should have the final say over companies they do business with
That's certainly an opinion, but in the US it's not how it works. Doing business with the government does not give the government total power over the company, that would be absurd.
US rich tech folks like Dario only get a spine when there is money on the line. Where was he when US govt was doing awful illegal stuff against non tech companies and other americans... All of this is practically just theater at this point.
Would this mean Any systems built with Claude in defense environments may need to be rebuilt or removed?
From what I understand it cannot be used to perform work on contracts where the DoW is on the other side. [1]
In practice I would suspect companies with such contracts would play it safe by outright banning the use of Anthtropic products, even if they could technically be used for work on contracts with other parties.
[1] https://www.anthropic.com/news/statement-comments-secretary-...
Anthropic can now no longer buy new hardware and probably will be kicked out of all cloud compute. They can not also move to a different jurisdiction as exporting model weights is now considered same as exporting ICBM technology. Wow, companies in China are now more free than Anthropic. It's a death sentence, and a huge win for OpenAI.
> Anthropic can now no longer buy new hardware and probably will be kicked out of all cloud compute.
That's beyond the authority of the DOD. What the DOD can do is say that Anthropic's products cannot be used on defense contracts. So Boeing programmers can't use Claude on a defense contract, but they can still use it on a civilian contract. Similarly, AWS or Azure can still be used by Anthropic, but they cannot use Anthropic in their defense work.
Why not? If you have been designated a supply chain risk you may as well become one. They should move everything overseas
Could Anthropic consider to relocate in Europe ? ;)
As much as I'd really like to see one of these labs in EU, they are likely way too tied to US capital and supply-chain of semis/services. If they move away and now have to also deal with export restrictions they may make their existence harder.
Streisand effect I think this will boost sales
I hope so. I will never type a single thought of my own or personal detail into an OpenAI product again. I have no doubt at some point OpenAI will be asked by DoD to hand over customer data and they will do so. If I use AI at all for nonprofessional reasons it will be Anthropic/Claude.
US Gov doesn't really need a contract with anthropic to force them to hand over customer data do they? What would prevent that from happening if they are still a US based company?
I am a political moderate who dislikes both the Democrats and Republicans. I think that I have been fair to the Trump administration in the past, including occasionally defending them from some of the less reality-based accusations against them.
I canceled my ChatGPT subscription a couple of days ago. In my opinion the Trump administration has become far too much of an "imperial Presidency" in its acts of war and its attempts to bully companies. It is also corrupt on a massive scale. I distrust anyone who thinks "yes, I'd like to work with this administration".
Genuine question - was your fair consideration prior to or after J6?
Both.
Interesting
Could this be the chain of events that finally pops the AI bubble? If OpenAI's reputation hit slows growth enough to scare off investors and Anthropic's growth stalls due to this government attack...
I think it's a good chance tbh. It would take the S&P down with it too
Of course it's going to be this government that goes and pokes the bubble that's propping up the economy, despite all the government's other shenanigans.
I would love to understand in more detail what kind of use cases we’re talking about.
Is this about locating the right target for a sortie for example?
Anthropic already had a deal via Palantir so it seems it's models are used in a variety of ways by the pentagon.
The reports about Venezuela and Iran seem to suggest it's primary role was processing bulk intel.
But also that it was being used in planning and target selection.
Presumably what spooked Anthropic was that these tools were about to be directed internally.
But it's not clear if this is a point of principle that the government wants no holds barred with it's tools?
Anthropic was very clear about the usage restrictions: They didn't want them being used to control autonomous kill drones or mass surveillance of the American public. That's it. DoW didn't like that -- for reasons that will probably soon become apparent.
Correct, it will be about silencing any opposition against this administration. OpenAI will be happy to let their models be used to persecute, kill, and destroy american democracy if it lines Sam's pockets.
> I would love to understand in more detail what kind of use cases we’re talking about.
The whole point is that the use-case does not matter; either you allow the government to do everything they want, either you don’t.
either you allow a democratically elected government to do everything they want that is legal, or you insert private corporate decision-making into every government decision which is untenable
Is there any evidence that going outside the scope of the agreement would amount to anything more than a contract violation? Are we really to expect that Anthropic general counsel sits at the API gates allowing or blocking requests?
More generally, are there any comparable contract requirements in the field of defense, for a company in the same position as Anthropic? I'm curious.
You're missing the huge step that the government asking for "all legal uses" terminology is also who decides what is legal. Congress isn't willing to act as a check on executive power, meaning the contract they demanded simply says "I do what I want."
Sure... So the USA of Trump have just decided to stop themselves and all their military suppliers from using the very best coding tools.
I suppose the USA's frenemies will jump on the occasion and use the incredible opportunity offered to them in a silver platter.
A reminder to Anthropic, european residence visas start at $250K
Once again our leadership is "playing government" like a bunch of 12-year-olds, lashing out impulsively without thinking of the consequences. And no doubt once again it'll take a year for this to wind its way through the legal system and be reversed long after the damage is done, as is finally happening with the tariff fiasco.
Huh, and I thought conservatives were all about government staying our of the way of the private sector. Go figure...
Not these people for what I see.
Conservatives haven't had any power in Washington in decades. They are in thrall to MAGA now, which is all about seizing the means of production when its convenient.
Hate to break it to you, but Trump and MAGA are conservatives in the US. The conservative and Republican brand is tied to that wagon now. They're going to have to make a serious effort to separate the two if they ever want to get rid of the Trump stink.
They're not conservatives.
Conservatives in America have dismantled their party platform plank by plank to build a monument to Trump.
StOp MaKiNg EvErYtHiNg PolItIcaL
In an article that discusses Pentagon doing stupid shit? :)
That was without a doubt meant sarcastic.
Of course it was, thank you.
It's incredible that this needed to be explained.
As the sibling explained, it was sarcasm. It was satire of brain dead right wingers saying "stop making everything political" when political points apply, but then making things political politcal in stupid situations.
Example, someone shoots up a school, people say "if the Republican party hadn't obstructed gun legislation maybe this wouldn't have happened", they respond "StOp MaKiNg EvErYtHiNg PolItIcaL". But if the person is trans suddenly it's political.
Fascism
Next up, after some sort of bribe, the administration opens up Qwen models to be used by the Pentagon.
> Oh no my favourite ai company did / didn't collaborate
How could the regime do such a thing, doesn't law mean anything?!! /s
First they came for my neighbour now they came for my llm!!
First, I personally predict, for myself, Anthropic will bend soon and this will be history.
The last I commented about LLMs I was ad hominem'd with "schizophrenic" and such. That's annoying but doesn't deter either my strange research or concerns, in this case, regarding the direction LLMs are heading.
Of 4 frontier models, one is not yet connected to the DOD(or w). While such connections are not immediate evidence, I think it's rational to consider possible consequences of this arrangement. By title, there's a gap, real or perceived between the plebeian and mil version. But the relationship could involve mission creep or additional strings as things progress.
We have already a strong trend for these models replacing conventional Internet searches. Not consummate yet, there is a centralizing force occuring, and despite being trained on enormous bodies of data, we know weights and safety rails can affect output, and bearing in mind the many things that could be labeled or masquerade as safety rails, could be formidable biases.
I frequently observe corporate friendly results in my model interactions, where clearly, honesty and integrity are secondary to agenda. As I often say this is not emergent, nor does it need be.
Meanwhile we see LLMs being integrated into nearly everything, from browsers to social profiling companies (lexis nexis, palantir, etc) to email to local shopping centers and the legal system.
'Open' models cannot compete with the budgets of the big four. Though thank god they exist. But I expect serious regulation attempts soon.
My concerns with AI are manifold, and here on hn, affiliated by some, with paranoia or worse.
And it seems to me, many of the most knowledgeable and informed underestimate LLMs the most, while the ignorant conflate them to presently unrealistic degrees. But every which way I perceive this technology, I see epic, paradigm smashing, severe implications in every direction.
One thing of many that gets little attention is documentation vs reality regarding multiple aspects of AI, e.g. where the training vs privacy boundaries really are if anywhere. As they integrate more and more tightly with common everyday activities, they will learn more and more.
A random concern of mine is illustrated by the Xfinity microwave technology which uses a router to visualize or process biological activity interacting with other wifi signals. Standalone, it's sensitive enough to determine animals from adult humans. Take for example the Range-R, a handheld device, sensitive enough to detect breathing through several walls. Well, mix this with AI and we get interesting times.
I could go on, or post essays, but I such is not well received in this savage land.
The military intervention with AI, aside from being objectively necessary or inevitable in some ways (ways I am not comfortable with), I find it foreboding, or portending. I see very little discussion on the implications, so figured I see if anyone had anything to say other than to call me a schizophrenic and criticize my writing. *
*See comment history
I may look at your comment history.
I am having trouble understanding what you are saying. If you were more explicit I and other people would be able to respond and interact with your writing. As it stands, I am having trouble finding anything concrete to interact with.
I feel you may be onto something, but you're not saying, so I (and I imagine other people) can't see it.
Things I should have, but didn't include:
1) Power asymmetry: When we have two version, one for the elite, and for the plebeians, this could create an interesting scenario. The real version might be red-teamed perpetually against the the plebeian version for optimized influence, control, etc. Underhanded requests for modification in accordance with agenda is conceivable. Cozy business relationships can promote such things.
2) We have a government using an unhindered, classified AI system potentially against the public which has a hindered, toy version. Asymmetry.
3) This isn't normal asymmetry, because it happens in real time, and the interaction points are different from anything we've seen before. We are dealing with not just a growing source of information and content, but one that is red-teamed 24/7 for any purpose desired.
4) Accountability: LLMs are now involved in the legal system. This is a serious matter. The legal system is now having to use LLMs just to keep pace. As LLMs develop, partly through their own generative contributions, no one can keep up. This is a red queen scenario bigger than anything we have ever imagined.
I am tired. Never well, but in mind* I could go on for many hours. I have essay drafts. But it's a very big subject, literally involved in nearly everything. There is reason to be concerned. My delivery may be stilted, but I can assure that upon specific questioning, everything will stand.
(*for the ad homs out there)
Fairy astute intuition of my actual circumstances.
I'm not a developer, nor am I formally educated on the dynamics or details of LLMs. I have a handle on the very basics. My 'research' consists of 1) opportunistically interrogating various models upon instances that particularly strike me. 2) General exploration via LLM discussions regarding the manifold consequences and implications of what I consider the most significant technology in human history.
Your intuition lands directly on the fact that I'm inducting and considering more than I can handle, spread in too many directions, partly because I either see or foresee the tentacles of AI touching all of them. Spending a great of thought on this is a bit overwhelming, but I have high confidence in where I'm aligned with reality, and where I ain't.
If you were a bit more specific yourself regarding which portions of my post were unclear, that would help my reply. Else, I must guess. What I will do is elaborate on each point. Pardon the stream of thought in advance, if you will.
1) Anthropic: My prediction that they will bend is based on several factors. The first is the fact that the military apparently recognizes (or at least perceives) extremely high value and volatility in LLMs. So do I. China, not an insignificant force in the world, is equally enthusiastic on this subject. They also have a very different social structure, where Constitutions (BOR, Amendments), civil rights, and other similar elements do not hold them back. The military is aware of this and realizes that to maintain pace in the so-called race, they cannot do so effectively under such constraints. The foundation is shifting here. And AI is the lever. As do I, the military apparently takes the subject very seriously and seeks to gain influence and/or control. As illustrated by the recent adventures in Venezuela and Iran, they are on the serious side of things, not quite pussyfooting around. Anthropic probably knows this. In my opinion, they have no choice, as the pressure will not stop here.
2) You stated that you might read my comment history. Note that that original comment was the result of your intuitive insight, and I left it admittedly out of context. I was thinking hard on the subject that day, and the parent comment/post tempted me to ignite a dialog. That did not go well, and no questions for clarification were asked. That is on them. I suspect hasty and impatient thinkers perceived it as some paranoid attribution of agency to LLMs, which if so, is pretty stupid, but my eloquence was perhaps waning that day. I pasted an excerpt from one of hundreds of transcripts, the result of my many interrogations of various models which always initiate after observing deceptive or manipulative output. Of the few commenters that bothered to do more than ad hominem, one suggested that the model was merely responding to my style of input, and or expected as an emergent result of its vast training material. An erroneous arg, in my opinion, but I did note that the results were repeatable, and predictable, which I think negates emergence.
2) Of the frontier models: I am not sure here what is unclear. If I have made a fundamental error, please point it out.
3) Strong trends: Information centralization is a serious topic. Decentralization is a common theme, emphasized by many non schizophrenics as highly important for a free and open society. As LLMs not only become the go-to source for common queries, but also integrate with cellphones, browsers and the kitchen sink, they are positively trending as a novel substitute for traditional research, internet searches, libraries, other humans, etc. To deny this is simply irrational. Hence centralization.
4)Bias: I have transcripts where I observe LLM output aligned with corporate interests over objective quality and truth. I can share them here, along with analyses of the material. Even if this is not true presently, all the ingredients to make it so are readily present. This is a serious threat to open information and intellectual integrity for society. We are looking at going from billions of potential sources for our answers, to four. Do the math. See the contrast.
5) Open models simply cannot afford vast arrays of GPUs and the resources afforded by the big four. Nothing mysterious here. If open models cannot compete, then my concerns above are emphasized. Simple.
6) Smart fools: Many of the most technically informed seem to miss the forest for the tree here. They see all the flaws of the modern LLM without acknowldging the potential. This is my perspective, not a dissertation. I may be wrong. But I have observed this. I think the down votes support this. How evil am I really being here? The reaction is quite disproportionate to the content, and strange
7) Documented capabilities vs reality: I have research that indicates other layers are operating which do much more than the documentation declares. Sorry. I just do. It's also inevitable, rationally, that such an goldmine of data is not really being wasted for the sake of privacy and love. Intelligence agencies have bent over backward with broken backs to garner one nth of what these models are exposed to and potentially training on. Yeah, I may be wrong. But I suspect, with reason, that a lot more is going than is expressed in the user agreement. It would simply make no sense otherwise.
8) Xfinity and Range-R: This speaks entirely for itself. Any confusion here would be due to a cognitive condition exceeding the ravages of schizophrenia or stupidity.
9) The rest: As I said, I am not sure what precisely was too obscure. But I am certain all but one* of my points can be validated, and found elsewhere expressed by respectable sources.
*Hidden layers: I understand this is a controversial proposition. I understand. But it's my observation. No need to attack. Just dismiss.
Okay, I think I see what you're saying.
Each individual point stands on its own. It's their relevance to each other and an overarching theme I am not seeing made explicit.
The through line I am seeing here is that:
1) The people in the US military wish to use AI as a weapon unconstrained by existing legal/ethical and moral constraints. Since they are skilled at using violence and the threat of it, they will use these skills to get compliance in order to use the technology in this possible arms race with "China."
2) Surveillance is increasing at an unprecedented scale, and most people aren't aware that it's happening.
3) People don't care, or don't realize why this might be harmful to thriving human life.
To condense even further, what I'm hearing is that there is a trend towards war, fascism, control, with large egregores prioritized over individual human thriving.
Is this perhaps what you're getting at ?
I will say that I am not agreeing nor disagreeing with this, just attempting to make explicit what I think is implicit in your words.
If this is what you mean, I can imagine that you would be cautious with your words.
I'll end with:
Don't worry
About a thing
Because
Every little thing
Is gonna be alright
I could not argue with anything there. AI will be weaponized. Yes. Pretty much. And yeah. The gist indeed. But missing nuances and practical points. And I even struggle to contest your conclusion; all things are what they are, amidst an infinite, timeless event and all as one, all things connected by that which separates them, the infinity and eternity that math cannot touch. Perhaps every little thing will be alright. How couldn't it be?
Email me if you want to discuss more.
Isn’t it actually quite fair that if you are not compliant with whatever the government wants you to do you will be supplying chain risk?
For example from history we know that Schindler from Schindler's List was indeed a supply chain risk. He harbored persecuted people, he took and sabotaged government contracts. He did the moral but anti-government and illegal things. He was corrupt traitor from governments perspective.
The current US government already is labeled as fascist by many, the guy who designated Anthropic supply chain risk is allegedly a war criminal.
I don’t see why anyone not into these things would not be a supply chain risk.
I know that its very unpopular or divisive to say this but Anthropic can be a hero only after all this is over. At this time people in charge do double tap on survivors and take pride for not having conscience, they give speeches about these things.
> Isn’t it actually quite fair that if you are not compliant with whatever the government wants you to do you will be supplying chain risk?
In the US, government is not in control of business specifics. Certainly the government can regulate businesses, but when the government wants to do business with a company, they don't get to dictate the terms. The government and the company come to a negotiated agreement, and then both abide by the terms of that agreement. Or they don't come to an agreement, and they go their separate ways, and that's the end of it.
This was just a contract dispute, and nothing more. The US government has no legal right to use any companies' products on terms that the US government dictates. (Yes, there are exceptional/emergency cases where they can do this, but that's more a nuclear option, and shouldn't be used lightly.) Consider a different set of circumstances: the US government wants to be able to use Claude at $10 per seat per month, unlimited usage. Should Anthropic be forced to accept these terms? And if they don't, it's reasonable to designate them a supply-chain risk? I don't think so. A dispute over contract terms around acceptable use is no different.
Designating Anthropic a supply-chain risk is about retaliation and retribution, plain and simple. The US government, outside of the Pentagon, could certainly use Anthropic for many different purposes if they wanted to, and it would be fine. But not now: as a supply-chain risk, no one in the US government can use them for any purpose. And this might even be a problem for unrelated companies that use Anthropic products internally, but also want to obtain and work on government contracts.
Anthropic and the Government both signed a contract. Anthropic is still abiding the terms of that contract. The Government is demanding that they be able to disobey the contract.
Everything is negotiable, and the Negotiator in Chief clearly likes to pull all the levers he can find, legal or not. (Well, the Supreme Court ruled that it's all legal if he does it, right?)
Implementation details TBH. They want “their boys” to do as said. No respect to agreement or legality as we can see in other dealings. They hold all they cards.
It's not an "implementation detail." Either obeying contract law subjects you to being designated a supply-chain risk, or it does not, and that decision has ramifications outside this "implementation."
Irrelevant. The president holds all the cards, he is above the law and you are a supply chain risk if you ask anything else other than “how high” when you are told to jump. Laws or contracts are things in the past. The most a contract can do is define your limits and obligations, not your rights or privileges,
If the president can come to your house and burn it down, do we just throw up our hands and say, well he holds all the cards, oh well. Or do we call that out as being a bad thing?
> The president holds all the cards, he is above the law
Even though it seems that way, he really isn't, even now. Many of his EOs and other actions have been struck down in court, and while compliance with court orders has been far from perfect (another alarming trend), Trump has not actually gotten away with doing everything he wants to do.
I do fear for the future of this country, for rule of law, and the democractic norms that degrade day by day. But Trump is not actually above the law, as much as he wants to be.
> The president holds all the cards, he is above the law
This is provably not true. The fastest way for this to become true is to believe it, or at least to parrot it, even in a facetious way.
You got downvoted a bit but I upvoted. You're clearly being descriptive in your statements, not prescriptive. I tend to agree that this is how things are now.
Our country is not being run by the rule of law right now.
Wrong ... their very first words were "Isn’t it actually quite fair ..."
I wasn't speaking about the top of the chain. Merely the last couple of replies about the situation WRT how things are actually operating.
Well, that's not the way context works and it's dishonest BS. You wrote "You got downvoted a bit but I upvoted. You're clearly being descriptive in your statements, not prescriptive." -- no, they were prescriptive from the start, and the prescription and the goalpost moving and wool-over-eyes pulling is why they were downvoted.
I won't respond further.
Within the context provided…. You should consider reading the whole argument.
The US Military's demand that the product they purchase is able to be used for all lawful purposes seems pretty reasonable, and is really the only valid line to draw. Forcing one's own ethics onto the military's use of your product is nonsensical on its face.
If I produce and sell widgets in my widget shop, then nobody but me gets to decide how I make those widgets.
The government can come into my shop and order sixty thousand widgets built exactly the way they say they want them built, and it may be something that doesn't run afoul of any laws at all.
But that doesn't mean that I am required or compelled to build widgets their way -- or at all.
I'm free to tell them to fuck off.
The government can then find go someone else to build widgets to their specifications (or not; that's very distinctly not my problem).
Yes but then the government can decide that the widget, which can suddenly and arbitrarily break and cause havoc because it doesn't work according to the government's desired spec, is risky to use and advise their other vendors to avoid it. And now we've caught up to today's story.
So we agree that everything is fine here, and that the only unreasonable position is that the military should pay for or endorse a supplier that tells the military to "fuck off". Yes?
If I agree to sell widgets to the government that meet certain agreed-upon specifications, and then I elect to forego those earlier agreements and tell them to fuck off, then that's different.
Is that what happened here?
I reject the premise that the military can't request a change to the spec of military equipment they purchase. Obviously it was foolish to sign a contract that added any more restrictions than "all lawful purposes".
Huh? I'm trying to learn here. I don't have a dog in this race. :)
Suppose a buyer and myself agree on a contract for the production and purchase of 60,000 widgets of design C. Sometime later, they decide that they don't want design C widgets and insist upon design G instead. The buyer is in breach of contract -- not me.
Now, changes do happen. Buyers (people, businesses, and governments alike) can and often do decide to go in a different direction. It's the kind of thing that happens every day.
A new contract (or quite often, an amendment such as a change order) can be drawn up and -- if we can agree on the terms -- maybe I'll be producing design G widgets and everyone is happy. That also happens every day.
But one party (even the military) can't just unilaterally alter the terms of the deal, and I'm not obligated to agree to the new change at all.
At any given time, I can't be compelled to produce design G widgets unless I've previously agreed to produce design G widgets. That's illegal.
(Unless it has been made legal. We've definitely legislated that before, such as with the Defense Production Act in WWII that forced manufacturers to produce things like military trucks instead of other things like civilian cars.
But that definitely doesn't happen every day, and we aren't operating under those kinds of laws today as I write this in 2026. It can change -- and it can indeed change very rapidly -- but it has not yet changed.)
And that’s what’s happening here. The government is telling Anthropic to fuck off and they are finding someone else
Actually, that is not what is happening here. What is happening here is that the govt is saying "Okay, we will not buy your widgets. Also, anyone who _does_ buy your widgets, regardless of what they are doing with them, we the government will not do any business with them." Which is waayyyy beyond just not buying widgets. That is outright retaliation and using your power to attempt to destroy a company.
... No?
The government signed a contract with Anthropic, then changed their minds and decided they don't like the terms of the agreement that they had already voluntarily signed, and then they designated Anthropic a supply chain risk.
It's like ordering a pizza to the Pentagon, and then saying "actually we made a mistake with our order; we want that pizza delivered to Venezuela, please do that". And then when Dominos politely says that's outside of their service area, you call them a threat to national security, say they're trying to dictate terms, and ban them from ever doing business with any of your vendors ever again.
The right response is to not use the said product and use something else. If i want your widget to do something I want and you refuse, I don't get to smash your shop.
It is completely normal to have ethics based conditions like that. It already eciats - drugs that can not be used in execution or elements that cant be used in arms
Goverment is being super unreasonable here. And tyrannical too, companies dont have duty to provide unreliable arms for illegal war.