I know everyone says this is political but it actually seems like a textbook designation. Anthropic wanted to have rules on how the military used AI, the military said no and therefore doesn't want anthropic used anywhere in their supply line.
This is like a pen manufacturer not wanting their pens used to sign drone strike orders, now the military needs to have a special box of pens that don't have stipulations attached. With AI usage it would be the same thing except applied to entire product chains. It seems like it would just add more complexity to operations.
You can agree with the rules anthropic wanted, but having rules set by a private company at all that apply to the military does seem fair for the military to object to.
>The Department reasonably feared that Anthropic might manipulate Claude’s design to prevent it from performing national-security functions that the Department deems contractually authorized and necessary
Though they'd probably put the DoD on the cybersecurity whitelist today, the very idea of the claude whitelists for certain functionality already exists and is being used by them today.
The reason it appears corrupt is that OpenAI has the exact same restrictions (https://openai.com/index/our-agreement-with-the-department-o...) but was not declared a supply chain risk. If two vendors have the same restrictions and they only designate one, then the designation must be arbitrary and/or capricious.
No it doesn't. Specifically OpenAI had a "all lawful purposes" clause (which plausibly includes the existing mass surveillance apparatus), but Anthropic did not.
I thought the difference is that clearly written rules or SLAs are acceptable and can even be negotiated before signing a contract, but having a person, in this case Amodei himself, manually approve DOJ's usage case by case with his own moral judgement after a contract is signed is not okay.
Do you think OpenAI's terms for the DoD are published on a website? Pretty sure when Anthropic got zapped, OpenAI sent DoD a memo 500ms later saying they wouldn't restrict use.
Look at it pragmatically. What does the DoD use every single procurement for? Hint: military use.
> No use of OpenAI technology for mass domestic surveillance.
> No use of OpenAI technology to direct autonomous weapons systems.
> No use of OpenAI technology for high-stakes automated decisions (e.g. systems such as “social credit”).
These are stronger than Anthropic's restrictions (https://www.anthropic.com/news/statement-department-of-war), as OpenAI says themselves: "We think our agreement has more guardrails than any previous agreement for classified AI deployments, including Anthropic’s. [...] Based on what we know, we believe our contract provides better guarantees and more responsible safeguards than earlier agreements, including Anthropic’s original contract."
The article only cites language for two paragraphs, which is the surveillance.
The other stuff...there's no citation or quote. Meaning that it could be OpenAI having their cake and eating it to, with no pushback on the actual language, or consequnces, or enforcement.
I think you're confused. When the White House decided to stop having federal agencies buy paper straws [1], they didn't designate paper straws a supply chain risk, they just stopped buying them. The term has a very specific meaning which would not apply in the pen scenario.
Non-sense. Pentagon has many many uses of AI besides autonomous weapons and mass domestic surveillance, which Anthropic doesn't restrict. ie: intelligence analysis, target selection, payroll, inventory, research, supply chain and logistics, medical, bureaucracy, ...
> This is like a pen manufacturer not wanting their pens used to sign drone strike order
I am pretty receptive to the "guns don't kill people, people kill people" argument, but we are taking a big leap from pen to llms/artificial intelligence. I am sympathetic to some of your other points, but I simply must reject this analogy.
Anthropic's red lines were 'no fully autonomous AI kill chains' and 'no domestic mass surveillance'. Those are clearly not national-security or DoD functions in the first place, so why would they be in any way objectionable?
OpenAI has hacked several prominent entities and is allowed to do business as usual but Anthropic putting guardrails on the military's usage of AI is the national security threat while Pentagon itself said that over reliance on AI lead to that attack on school in Minab.
I have also learned that OpenAI's president is a big maga donor and now I am thinking of cancelling my OpenAI membership todau and signing up with Anthropic again.
Edit - i am not a OAI hater and i don't hate anyone. I am just saying Anthropic is kinda getting bullied and OAI is not being punished enough for their actions. It is time for me to support Anthropic (with my $$$) instead of OAI.
>OpenAI has hacked several prominent entities and is allowed to do business as usual but Anthropic putting guardrails on the military's usage of AI is the national security threat while Pentagon itself said that over reliance on AI lead to that attack on school in Minab.
Right, because "supply chain risk" designation is as it pertains to them being a supplier to the military, not a overall assessment of their standing as a corporate citizen. Not to mention that the administration wants to go full steam ahead when it comes to AI development, so there's no internal contradiction here.
These are just very different things. OAI "hacking" anything isn't OAI telling the US Gov what they can and cannot do with their product. Maybe along all kinds of axes, one is more alarming than the other, but it's a canard in this topic.
>> These are just very different things.OAI "hacking" anything isn't OAI telling the US Gov what they can and cannot do with their product.
OAI hacking the chinese government will definitely raise alarms in DC and will be bad for national security. Just because the country that was hacked (australia) is a weaker one doesn't mean the underlying action is not troubling.
>OAI hacking the chinese government will definitely raise alarms in DC and will be bad for national security.
That seems like it's too easy to abuse as a cudgel for whatever companies you vaguely hate. Is ExxonMobil a "supply chain risk" because the effects of climate change might cause geopolitical ructions in the next decade?
Hate is a pretty strong word here. I don't hate OAI. I am a paying subscriber. All I am saying is maybe its time to switch, show support for Anthropic because it is kinda getting bullied right now and the OAI is not getting enough punishment for their illicit actions. Is it too hard to understand?
So you claim: If a similar thing would be filed against OpenAI, it would stand up in court.
That is still not corruption, yet. Also in Germany it is the executive branch which controls who is sued or policed by the stately organs.
So favorism can be passed down this way, but not we are not at corruption yet. If reciprocity is expected, then you could call it that. Or could perhaps even file for negligence in treating all other company actors the same, especially if you are Anthropic yourself.
Still, I blame the US voters as much as the current administration if the later is just clowns and thieves.
Also, IANAL and HN discussing US things is always foreign to me.
If the president favors one party over another with no clear benefit to the voters, that is very indicative of corruption. If his organization is also getting donations from said party, it is even more certain to be corruption.
You argue that the courts are not corrupt, hence there is no corruption. Without considering the president.
You're way overthinking this. The administration told Anthropic to let them do something, Anthropic said no, and then they got pissy and started throwing rocks and wouldn't back down because the defining trait of this administration is that they walk into stupid positions and then refuse to back down because backing down would make them look weak and wrong.
They literally lined up to shower trump with gifts after the inauguration. Tim Cook gave trump a 24K gold bar with an easily detachable plaque so as to not have just given him a gold bar.
Not just "a big MAGA donor," the biggest single contributor. Greg Brockman donated $25 million.
I get downvoted every time I point out that this was entirely rational. This, downvoters, is why it was entirely rational. Does that mean I agree with it? Of course not. But investments are made to pay off, and this one certainly has for OpenAI.
Actually, on further research, what I have wrong here is that Brockman was the largest single donor to the MAGA PAC after the election, contributing $25 million in September 2025 [1].
Yeah, it surprised me when I found out. I can intellectually understand why average people choose MAGA and try and be somewhat understanding even if I find it distasteful. My conclusion is that if you aren't personally worth at least $5M or have the direct phone number of someone who can get a Trump admin official on the phone, then it was probably a bad choice.
Of those who meet that criteria, and especially those with MUCH more money and closer connections, it actually is an extremely smart choice. However, it represents a level of reprehensible naked self-dealing and complicity that I simply can't condone. It really is comparable to the looting if public goods and coffers during and after the fall of the USSR. Nationalism and patriotism are dead.
Personally I am enjoying Muse the most and so far havent hit any limits. I also (i know it's not a popular thing) have enjoyed wearing my Meta Ray Ban's since 2023 even if they break as MANY months later they come back to life.
EDIT: Oh damn getting downvoted. It looks like Meta Glasses are hated as MAGA, maybe Zuckerberg should call them MAGA glasses ha
you are so honourable and ethical about this issue .. I really respect your courage in cancelling (or planning to) OAI, especially since you must be so used to the harnesses .. workflows .. and muscle memory youve developed around their agents .. and now you'll have to switch to Anthropic and relearn its patterns .. I really think there's a lot of dignity and courage in that .. and others could learn from you ..
I can 100% assume you showed that same courage and moral fortitude by quitting X after Elon acquired it. Respect!
Thanks for the encouragement. It's hard to come by these days on the internet. I never really used twitter even before elon bought it. I don't use any social media, maybe except hanging out on HN often.
I wonder if this is going start being abused soon.
If I was Palantir or any other GOP aligned company, I would be completely against this. What's to stop a Democratic president from doing the same thing and destroying them?
Because the modern Democrats are primarily a corporate dominated party. They haven't really had any hard hitting legislation nor legal regulation/rulings on corporations in easily the last 30 years.
I doubt that is going to change anytime soon. They are a private entity, so they can manage and change their primary system any way they want.
You got that exactly backward. Modern republicans are the ones that get the most in corporate PAC money, at the high end it's like 60/40% split... the majority to the GOP. Also, the fact that modern republicans brought Citizens United and just earlier this year NRSC v. FEC.
You really can't make a valid case to say the dems are the corporate dominated party. ffs, they are the only ones even attempting to put common sense regulations on the financial industry, something Trump called "bad for business" when he undid antitrust legislature that got passed under Biden.
I would argue the democrats are more a process oriented party. They wouldn’t pull this shit because they care about process - so much so that they wouldn’t even consider half the stuff that’s going on now as it circumvents process.
That's not responsive to OP or coherent at all really, even if rhetorically structured like a gotcha.
Incumbent Presidents in the US almost always run for reelection to a second term, that's the normal expected outcome with vanishingly few exceptions in the last century.
Exactly. Their process prioritizes seniority generally, and this is another case. The process is, if the president wants another go, he gets another go; and he sure as heck wanted another go.
And all the shenanigans around 2016 when Bernie Sanders was shut out by the DNC. Even Elizabeth Warren said outright that the DNC rigged the 2016 nomination. I think the outcome would have been very, very different if the DNC didn't fuck with the timeline and forced Hillary as the Dem nominee.
> Democrats are primarily a corporate dominated party. They haven't really had any hard hitting legislation nor legal regulation/rulings on corporations in easily the last 30 years.
Sarbanes-Oxley? Dodd-Frank? ACA? Now I'm sure you (like, every other living human) has a ton of nitpicks and criticism of bills like that. But they're real and they passed and they're largely-to-wholely partisan creations of the Democratic policy apparatus (SOX looked very bipartisan and was passed under Bush, but the actual bill was written mostly by democratic staffers).
I mean, fine, you'd like different regulation. But they're responsible for basically all the corporate regulation active in the US regime.
Bothsidesism is a very poor tool for this particular argument. Work with your allies, please.
"You must sell insurance to sick people!", "I don't wanna!", "I don't care, the law now says you must."
I mean, the whole insurance industry had spent the better part of a century (basically its whole existence) denying coverage to people with pre-existing conditions, and the ACA fixed it via fiat.
Not only is it clearly "hard-hitting", I'd argue it's probably the single most "hard-hitting" invasive change to corporate regulation of the post-war era. Certainly it is in terms of the amount of money involved in the affected market.
> how’s this anti-corporate?
And this is where the rub is. You're not saying that your desired regulatory structure is something different than existing proposals, you're demanding that your government engage in behavior that is expressly "anti" whatever you perceive "corporate" interests to be.
That's not democracy, that's just warfare. You have enemies and you want to win. And the problem there is that the structure of our government requires compromise.
The "everyone has access to the ER" has to be paid somehow, which is through tax dollars. The USA has a for-profit health system but typical neo-liberalism has a safety net where corporates get to dip into a different bottom-less pot. So it makes sense that we make this visible by taxing everybody who isn't paying their own insurance, since they still expect access to ER.
It "literally" did not. What enforcement exists is financial (a tax penalty) and not compulsory. And obviously most people are covered via employer plans or Medicare, as has always been the case. A quick Google says that 19.2 Americans are enrolled in ACA plans. I don't have statistics, but I'd bet that the overwhelming majority of those are willing purchasers. Most people view health care as desirable.
>The penalty for not having an ACA-compliant plan or forgoing coverage in 2018 is the greater of $695 per adult and $347.50 per child for a maximum of $2,085 or 2.5 percent of adjusted household income. The percentage penalty is capped at an amount equal to the national average premium for a bronze plan, the minimum coverage available by law
Ah yes, the very hard hitting anti-business rule that says if you don't become a customer of my friend's company, you must give me money. I'm sure the businesses were quivering in their boots when every citizen either became a customer, paid a fine, or became a felon.
Then why do the 7 publicly listed health insurers have shitty shareholder returns?
If their profits are so good, their shares would be worth owning.
Could it be because their profit margins are actually terrible (3% if lucky), and they have to get the price of their product approved by 50 different state government employees, and the federal government constantly changes their customer’s subsidies making revenue and expenses extremely volatile and difficult to forecast?
If you bought shares of those 7 companies in 2010 when the ACA passed, how much would you have today? More or less than if you bought the S&P 500?
If you're looking for policies to blame for insurance companies having a rough 2025-2026, look to the Republican-led efforts cap Medicare fees and alter eligibility requirements.
You would have about the same or less. Maybe UNH gets you a couple extra percentage points per year, but that's because they have an enormous healthcare provider business instead of just being an insurer. But it's still nowhere near a positive risk adjusted return when comparing to a relatively risk-less investment in SP500.
Since Jan 1, 2010, 18% for UNH, 15% for MOH/CNC, 14% for CI/ELV/HUM/SP500, 8.5% for CVS.
Also, these businesses are competing to sell a completely fungible commodity good with tons of non profits like Kaiser Permanente, HCSC, Cambia, Independence Blue Cross, and many others. It makes no sense to expect those businesses to be able to earn any decent profit margin when their competition only needs 0%.
Seeing Dodd-frank in the list is pretty hilarious, because it was put in place for the issues that arose from repealing Glass-Steagall by Clinton. Which arguably was one of the biggest causes of allowing the activities that lead to the Great financial crisis in 08.
And it is true, that it was signed by Clinton. But it was otherwise an entirely partisan bill passed with exclusively republican votes. All 44 democrats voted nay.
Now, given that, do you really feel that that's a supporting argument to the upthread contention that Democrats are a corporate-dominated party and not interested in regulation?
Please point to the hard-hitting ones, almost all the big co's at the time thought they got off too easy and were expecting a lot worse. At the end of the day, Sarbox and Dodd-frank basically just required those businesses to increase headcount 10%.
the cold political war turned hot with the prosecution of Trump (he made it easy by being an idiot). there will be no considerations for retaliation now, both sides will be using whatever means they have available to damage each other.
also if you look at how the two factions separate from each other decade by decade (there used to be overlap, almost none now) it's obvious that there will be no reconciliation.
it's a musical chairs game at this point, whoever is in power when the democratic facade cracks will be able to seize power and transition into whatever single party state follows.
What's to stop indeed? Some of us in the tech industry want the dems to slice and dice every tech company into irrelevance. They've betrayed society and they should be rightfully punished in fun + creative ways.
Not just Anthropic, but practically speaking, the leading foundational model labs like Anthropic and OpenAI are burning through enormous amounts of money, and will need to pay that money back. If they are designated as supply chain risks, or controlled technologies, their ability to achieve that ROI is severely constrained.
They are not turning a profit, and aside from enterprise token billing, every other offering they have is a loss leader. In order to flip that around they need to drive down the cost of inference significantly, either through massive compute improvements, more efficient or lower cost models, and either one lowers the barrier to entry for local and/or private cloud inference for open models. The moat established by the expertise in model training is effectively dissolving as they approach recursive self improvement because any mechanism that prevents model distillation effectively reduces the quality and utility of the models, so each improved model makes their competitors better by default if they want to realize those gains.
The reason for the pivot to control and regulation seems to be more about erecting an artificial moat in their market, and having these technologies controlled or highlighted as supply chain risks will continue to drive investments in alternatives, especially in other countries where data sovereignty and domestic tech has become a priority given the last decade of the US dominated tech industry.
I don't think that AI is going to go away, but I think the future of AI in the next several years looks alot like massive overfunded foundational model providers collapsing, their employees launching dozens or hundreds of direct competitors, and the leading hyperscalers buying up datacenter capacity the way they bought up dark fiber 20 years ago.
Can you clarify what a GOP aligned company is? Is that any company thats willing to do business with the U.S. government or any businesses thats willing to sell products in red states?
Maybe it's the companies that flank the US president at events in an attempt to gain favor? Or the companies that own private prisons that lobby for immoral policies on policing?
This doesn't appear to be unique to the GOP. Plenty of companies flank to whomever is in office. Even at that local government level. "Immoral policies" is subjective and would be in the eye of the beholder. If this was truely quantiative the U.S. wouldn't be a divided nation.
I don't see a need to come up with a complete definition at this point. Let's start with cancelling all SpaceX and Palantir contracts and see how we're feeling; that may be enough to keep corporations in line.
Because there is none. Its rhetoric and its personal to you. Its like me saying I haven't spent a nickle on anything Nike because of their narative selling. Hence anyone who wears Nike products is a bad person. Its ridiculous, narrow minded and undemocratic.
Telling my mind about anyone or any company is the ideal expression of free speech and democracy. Using government force against a company that refused a contract is the opposite of democracy.
It's certainly rhetoric! Corporate leaders must be made to understand that the business risk of collaborating with fascism is greater than or equal to the business risk of standing up for the rule of law.
I still don't fully understand what this was about. It sounds like the DoD said they wanted unrestricted access to Anthropic's models, Anthropic refused, so the Pentagon declared that they wouldn't use them at all.
As I understand it cane down to one point: Anthropic wanted the ability to review actions that would have an effect on human life.
Which sounds sensible on the surface, until you realize that all military organizations have extensive experience with that exact question, and have spent centuries refining their practices. Deferring that to a private company is actually a major regression and a concerning role to pass on to the public regardless.
FWIW I am not pro military, but it's a prominent part of human culture and not going away anytime soon. I do understand the nature of professions, however, and only one of these groups has made the interpretation of this question their specialization.
You raise an interesting point. Although wouldn’t Anthropic be the best experts on when a model would be most applicable or likely to falter?
I guess my opinion would really depend on the exact wording and details of the arrangement. Lots of people jumping in with strong stances based on pretty limited details.
On faltering: the US Military already has rules around the use of automated systems for this type decision making. This is not a new concept to them by any means, and currently it always involves a human decision.
> [Anthropic] would not permit its technology to power fully autonomous weapons without human oversight over targeting and firing decisions. Defense Secretary Pete Hegseth argued that the Pentagon should have unrestricted access to AI systems for “any lawful purpose” and that a private contractor should not be permitted to dictate how the military can use its technology.
Evidentially not - Anthropic want to do business with the US military, but on their own terms, telling the military what they can and cannot use it for.
I don't care what the military does. This is pure thuggery and abuse of the rights of private firms. Any time you contract with the government and they don't like your terms, will they designate you a threat?
In this context, that designation just means that the DOD won’t use your product and it’s suppliers can’t use your product for DOD-facing activity. That’s not “thuggery” it’s something any buyer can do. Walmart or whatever could do it too.
It’s not a problem. But it’s also not a problem for the DOD to decide that it doesn’t want to use anthropic or have its suppliers use anthropic for DOD matters. The DOD didn’t ban private citizens from using Anthropic. You can sell in your own terms but the DOD can also buy in its own terms.
That is the funny thing to me. I hate diving into politics here on HN, but this has such a rank smell to it. Trump's comments about Dario. His insane propaganda video about how we will never be communist / Marxist. In the same breath our government is bribing people with offers of $$ (the $5,000 checks) and essentially bullying to outright moving towards saying this technology is so powerful we will just have to seize it for our own use.
It is impossible to miss the absolute hypocrisy of this moment and how people and corporations are supposedly free, but if you happen to be of interest to the government they will try to casually ruin you. It has always been this way, people just haven't noticed because it flies under the radar most of the time when the govern crushes some smaller enterprise or another because it got in their way or didn't play by their rules.
It seems simple to me. Anthropic is a private company. If you don't like it fix the laws or don't buy from them.
Yeah, for national security being the biggest one. Also Congress can do whatever they want if they desired. If Congress mandated that all social media companies have to become workplace democracies, that's totally fine and legal.
We've forgotten what it means to run a society, we're allowed to change rules how we see fit whenever we want to. Only the elites exercise this power, but people have it too.
It's not good for the people. Anything that erodes the rights of companies erodes your rights. Private property and being able to run your firm on your own terms applies to everyone. Governments eroding them show they can do the same to you willy nilly.
> Private property and being able to run your firm on your own terms applies to everyone.
In theory, maybe. In reality, this is a privilege for a very small number of rich people.
> Anything that erodes the rights of companies erodes your rights.
How absurd. Food safety regulations don't erode my rights one bit, but they do erode the rights of companies to negligently poison me. I guess I'm losing the right to start a food company that poisons people, but I'm happy to forfeit that right in exchange for not being poisoned. We could extend this list of reasonable tradeoffs in which we all exchange the right to treat people poorly for the right not be treated poorly, and we all would benefit.
Except the people who want to sell poisoned food, those people will be worse off. I am okay with that.
Every private citizen or corporation gets to do business (or not) with the military on their own terms. This is the difference between totalitarianism and a free country.
Anthropic’s position is also known as “having terms of service”.
To be clear the existing, signed ToS terms that Trump balked at were:
1. No mass surveillance of US civilians - illegal but we have all read the Snowden papers.
2. Fully unmanned kill chains on autonomous weapons - Anthropic have even stated they will support this eventually, they just didn’t think their models were ready and didn’t want the obvious, inevitable blowback.
The DoD is within its rights to cancel the contract if it doesn’t like the terms that it previously signed.
The concern, I think, in its most generous terms, is that they can't buy Anthropic's service the way the could, say, a wrench, because Anthropic can decide at any point to disable the military's ability (even transitively, through other vendors) to use their models if they determined that they do not like what the military is doing with that or if it violates their interpretation of the terms of service.
No it's not. Anthropic didn't want an AI to make battlefield decision on who gets killed because it's too unreliable. That doesn't mean they don't want AI used in the military supply chain.
And it also doesn't mean that Anthropic is supply chain risk merely for not wanting AI to be the active decision maker in live or death situations in war.
And Anthropic's objection was recently confirmed: the US military blew up a school in Iran due to an AI error, an killed more than a hundred kids.
Dario wanted to sell a $200 million chat bot to the US military and also dictate how it could be used. The government said no. And because of that Claude also cannot be deployed to the GenAI.mil AI infrastructure. This means that any sensitive work for the DoD that uses Claude will go through Anthropic's infrastructure, not the US military-hosted backend.
So yes obviously defense contractors are forbidden to use it.
If the government wanted to put iPhones in missiles, and apple said "we won't sell you iPhones to put in missiles" would the correct government response be to say "no government contractor can ever use an iPhone".
They are not "just a purchaser" though. It's the world's most powerful organization that is funded by the people and is supposed to be for the people. I'm all for businesses making FAFO decisions, but expect a lot more here.
The DoD and US military does not allow contractors to use iPhones or anything else that is not backed by government-hosted or compliant infrastructure, such as GovCloud. The fact that the Pentagon and Anthropic could not come to an agreement on the usage and deployment to GenAI.mil means that it can't be used.
The reasons why are incidental to that requirement for defense contractors.
> This also means that no companies that work with th DoD can use anthropic models for any purpose.
No, it does not mean that. I don't know why people have been making this claim. If you work for a company, let's call it X, and your work is not DOD related, let's say you work on a social media service, then X has no reason (from this designation) to prevent you from using Anthropic, it will not impact their defense contracts at all.
People have been making this claim because the Secretary of Defense personally issued a statement demanding that "no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic". We're not going to play this absurd game where government officials issue fascist instructions and we all pretend they said something more reasonable instead.
Except you and many others are accepting what he said as if it were true, that statement is false. And hiddencost and others are just parroting it instead of describing the real situation.
This designation means Athropic's systems can't be used in DOD work, but many companies that work with the DOD also do work outside that. That separate work is not restricted with respect to the use of Anthropic's systems. Repeating a lie as if it were true is just dumb, it helps spread that lie.
Again, it's unreasonable to play this game. The regime must be held to account for the actual things that they say, not whatever sanewashed policies we can backfill underneath them. If Hegseth's statements exceed his legal authority, that doesn't mean he's lying, it means that the government is knowingly pursuing an illegal policy.
If you don't like that, and would prefer to discuss a lawful supply chain risk designation, you should convince him to retract the statements or impeach him.
Anthropic isn’t aligned with MAGA political ideology so they found an excuse to abuse power against them, more or less. The excuse was because Anthropic prohibits the use of their models for autonomous killing without a human in the loop they were declared a threat to national security on par with Iran or North Korea from a supply chain perspective in the military. OpenAI then rushed in and volunteered to build the slaughterbots ASAP, likely specially tuned to whomever Trump has taken a disliking to in his ambien stupors. Welcome to the future!
How is it delusion? If they don't want to work with certain types of government contracts they have the right to state so. Its just thuggery from Trump.
Yes. That's pretty much how it works for fascism. The government uses powers meant for a legitimate purpose to instead intimidate people and the mechanisms of government meant to restrain such abuse fail or are taken over entirely. Have you ever studied history?
Especially since Anthropic themselves have been literally saying that AI might destroy humanity. Why is it a surprise when people take them at their word and designate them a risk to critical infrastructure?
Side note, the Anthropic bots are astroturfing the comments.
>Side note, the Anthropic bots are astroturfing the comments.
if this is more than "people i disagree with are commenting", you should be messaging hn@ycombinator.com rather than writing it as a side note in your comment.
He's probably a maga or adjacent who is pissed Anthropic didn't bend the knee. You don't need to be a shill to hate this horrid abuse of government powers against a private firm.
Anyone who doesn't agree with me must be MAGA, right? You think people who don't like Anthropic are automatically Trump supporters? I have personally experienced how evil Claude is compared with the other options.
personally, i think that anthropic is a supply chain risk. if the ceo writes 'pace the frontier' and wants to be in charge of the product, it's very risky (to the customer).
by analogy it would be like the us military renting all m4 carbines from a political party.
anthropic treat customers very badly. they have claude lecture the user about the moral law according to anthropic. they change the product all the time. they have safety refusals to degrade functionality, they have mandatory logging so they spy on how it is used.
it's unusable for a serious application.
on top of that it is unsurprising that a big tyrant does not take well to a little tyrant.
How is this a debate? Anthropic's AI was used in military/defense operations. When Anthropic discovered this and disclosed it publicly, they attempted to revoke access and went after the government demanding it not get used in such a way. This would be a supply chain risk to any government.
If Lockheed Martin disabled weapons systems when it discovered the Air Force using its fighters over Iranian airspace based on Lockheed's ideological views, I'm pretty sure they would be labeled a supply chain risk at minimum. What is the issue here?
The government canceling that specific contract with Anthropic is fine. It is too vast a scope to deny Anthropic to work with any other companies who also happen to be defense contractors. Lockheed Martin for example cannot be sure now if it won't be sued if some employee uses Anthropic for help with accounting or anything unrelated to defense.
Thats too vast a scope. Whats not to say they start enforcing bans on Anthropic in any branch or usage of Anthropic of your company if your company has entered into a defense contract with the government for something?
they should ban it, defense contractors shouldn't be using it XD.
fable 5.1 doesn't have any option to stop anthropic spying on it. you can't have that...anthropic contractors and actually the rationalist ea clowns themselves don't have security clearance
For all the downsides of losing the business of government contractors, it really would hold Anthropic to their stated values: they’ll never develop AI controlled weapons on this blacklist.
I don’t agree with the decision, but I don’t think courts should be able to easily overturn the federal government’s findings when it comes to national security in a specific decision like this.
Supply chain risk seems a massive over-reaction, but understandable that the government doesn't want to commit to Anthropic controlled software for any critical military application if they can't 100% rely on Anthropic support in being able to use it as they see fit.
Every government on Earth has legislation that allows it to force companies (and citizens) to do its bidding under certain circumstances. That’s pretty much the definition of a government.
The point is this not being the first tool in the toolbox. Other competitors have shown they are more than willing to build weapons of death (OpenAI) so as the government you put them as the preferred contractor.
What the orange turd is going is rather petty bullshit in a market where many options exist.
> understandable that the government doesn't want to commit to Anthropic controlled software
But that's not at issue here. Obviously everyone agrees that the Pentagon can issue RFQ's to whatever specifications it wants. They can buy whatever they want, for the same reason that you or I are free to sell whatever we want.
The subject at hand is an attempt to use national security regulatory power to coerce Anthropic into selling a product it doesn't want to sell.
> The subject at hand is an attempt to use national security regulatory power to coerce Anthropic into selling a product it doesn't want to sell.
It would be naive to think that a private company could deny the government access to any tech if they want it for military or surveillance etc usage. In this case the government isn't forcing Anthropic to do it, they have alternatives and are perhaps making an example of Anthropic for refusing.
You could even make a case that the "supply chain risk" designation DOES make sense. If Anthropic are opposed to certain uses of their software, then doesn't that potentially impact companies that are part of the supply chain as well?
Admittedly I still need to do some reading on the current ruling, but how this isn't seen as an open and shut case of vindictive prosecution (if that's even a thing for this branch of law?) is astounding. The administration couldn't have made it more obvious there was no merit to their designation of Anthropic.
Yes, anyone and everyone should have a right to refuse contracts with the government. Is that difficult to understand? If you find it difficult to understand then perhaps democracy and capitalism isn't for you.
Maybe, but only in a very idealistic, unrealistic, sense.
Could you imagine a hypothetical where we're at war and a private company has a potentially war-ending tech (nuclear weapon say) that they refuse to sell to the government?
If you can only make such absurd examples then you have failed. Paying taxes are not a contract between me and the government. They are a forced fee extracted of me. Everyone should be worried about the government threatening to issue supply chain risks if they don't like you rejecting a contract with them.
Opinion by Katsas and Rao. For those that don’t follow this regularly, those are two of the biggest Trump hacks on the bench.
I suspect this will be reversed en banc.
Anthropic’s argument is really clear and easy to understand: we can’t simultaneously be a supply chain risk and also be a company the DoD demands we change our policies so they can use our models without restrictions.
I suspect some people will reflexively downvote you for bringing politics into the matter, but yeah, this is like a textbook bad draw for Anthropic. Unless you're a lawyer for Anthropic, I wouldn't give this ruling much weight at all.
I agree with its designation. The security scans from Claude Mythos have been highly suspicious, wrong, and with a significant false positive rate. Claude also directly causes conflict within teams through excessive utility of "git blame" which weaponizes everyone against each other until only Claude has control. I've had very bad experiences with Claude and only Claude.
> Claude also directly causes conflict within teams through excessive utility of "git blame"
Lol wtf, do you think nobody here knows how to use git? this is like a Fox News comment, you are not a developer and are trying to spread fear because a word was used jokingly. "git blame" is just a diff tool.
This is like a pen manufacturer not wanting their pens used to sign drone strike orders, now the military needs to have a special box of pens that don't have stipulations attached. With AI usage it would be the same thing except applied to entire product chains. It seems like it would just add more complexity to operations.
You can agree with the rules anthropic wanted, but having rules set by a private company at all that apply to the military does seem fair for the military to object to.
>The Department reasonably feared that Anthropic might manipulate Claude’s design to prevent it from performing national-security functions that the Department deems contractually authorized and necessary
Though they'd probably put the DoD on the cybersecurity whitelist today, the very idea of the claude whitelists for certain functionality already exists and is being used by them today.
No it doesn't. Specifically OpenAI had a "all lawful purposes" clause (which plausibly includes the existing mass surveillance apparatus), but Anthropic did not.
Look at it pragmatically. What does the DoD use every single procurement for? Hint: military use.
OpenAI just said no surveillance of U.S. persons.
Anthropic could have gone further; I think it's likely they did.
> No use of OpenAI technology for mass domestic surveillance.
> No use of OpenAI technology to direct autonomous weapons systems.
> No use of OpenAI technology for high-stakes automated decisions (e.g. systems such as “social credit”).
These are stronger than Anthropic's restrictions (https://www.anthropic.com/news/statement-department-of-war), as OpenAI says themselves: "We think our agreement has more guardrails than any previous agreement for classified AI deployments, including Anthropic’s. [...] Based on what we know, we believe our contract provides better guarantees and more responsible safeguards than earlier agreements, including Anthropic’s original contract."
The other stuff...there's no citation or quote. Meaning that it could be OpenAI having their cake and eating it to, with no pushback on the actual language, or consequnces, or enforcement.
No, they want to have rules on how the military used Anthropic AI.
Military would still be free to use OpenAI for the nasty stuff
[1] https://www.whitehouse.gov/presidential-actions/2025/02/endi...
I am pretty receptive to the "guns don't kill people, people kill people" argument, but we are taking a big leap from pen to llms/artificial intelligence. I am sympathetic to some of your other points, but I simply must reject this analogy.
however anthropic also believes it is some kind of entity unto itself so it is not at all reliable.
OpenAI has hacked several prominent entities and is allowed to do business as usual but Anthropic putting guardrails on the military's usage of AI is the national security threat while Pentagon itself said that over reliance on AI lead to that attack on school in Minab.
I have also learned that OpenAI's president is a big maga donor and now I am thinking of cancelling my OpenAI membership todau and signing up with Anthropic again.
Edit - i am not a OAI hater and i don't hate anyone. I am just saying Anthropic is kinda getting bullied and OAI is not being punished enough for their actions. It is time for me to support Anthropic (with my $$$) instead of OAI.
Right, because "supply chain risk" designation is as it pertains to them being a supplier to the military, not a overall assessment of their standing as a corporate citizen. Not to mention that the administration wants to go full steam ahead when it comes to AI development, so there's no internal contradiction here.
OAI hacking the chinese government will definitely raise alarms in DC and will be bad for national security. Just because the country that was hacked (australia) is a weaker one doesn't mean the underlying action is not troubling.
That seems like it's too easy to abuse as a cudgel for whatever companies you vaguely hate. Is ExxonMobil a "supply chain risk" because the effects of climate change might cause geopolitical ructions in the next decade?
That is still not corruption, yet. Also in Germany it is the executive branch which controls who is sued or policed by the stately organs.
So favorism can be passed down this way, but not we are not at corruption yet. If reciprocity is expected, then you could call it that. Or could perhaps even file for negligence in treating all other company actors the same, especially if you are Anthropic yourself.
Still, I blame the US voters as much as the current administration if the later is just clowns and thieves.
Also, IANAL and HN discussing US things is always foreign to me.
You argue that the courts are not corrupt, hence there is no corruption. Without considering the president.
If you find a person/company/agent odious, why would you want to financially support them?
You're ascribing complex motives to toddlers.
(Hmm did you edit your reply?)
https://www.washingtonpost.com/wp-apps/imrs.php?src=https%3A...
I get downvoted every time I point out that this was entirely rational. This, downvoters, is why it was entirely rational. Does that mean I agree with it? Of course not. But investments are made to pay off, and this one certainly has for OpenAI.
Which if anything seems worse.
1: https://www.theverge.com/ai-artificial-intelligence/867947/o...
Of those who meet that criteria, and especially those with MUCH more money and closer connections, it actually is an extremely smart choice. However, it represents a level of reprehensible naked self-dealing and complicity that I simply can't condone. It really is comparable to the looting if public goods and coffers during and after the fall of the USSR. Nationalism and patriotism are dead.
EDIT: Oh damn getting downvoted. It looks like Meta Glasses are hated as MAGA, maybe Zuckerberg should call them MAGA glasses ha
I can 100% assume you showed that same courage and moral fortitude by quitting X after Elon acquired it. Respect!
If I was Palantir or any other GOP aligned company, I would be completely against this. What's to stop a Democratic president from doing the same thing and destroying them?
I doubt that is going to change anytime soon. They are a private entity, so they can manage and change their primary system any way they want.
I suspect this is changing. The DSA branch especially wants all contracts with Palantir eliminated.
You really can't make a valid case to say the dems are the corporate dominated party. ffs, they are the only ones even attempting to put common sense regulations on the financial industry, something Trump called "bad for business" when he undid antitrust legislature that got passed under Biden.
* removed filibuster for lower-court nominations ("nuclear option")
* impeached the president twice in a term
* tried that former president in the Senate after his term ended
* repeatedly attempted to eliminate 60-vote rule for legislation
* expanded mail-in voting en masse
I'm not claiming these were illegal or not, just fast-and-loose with process/precedent is accepted. The process adapts to the needs.
We, the people, are on the other side.
This is such a tired argument that over simplifies an entire group of people on both sides into a single entity that happens to be the same.
Oh? Thats why Biden tried to run in 2024? Because democrats procedurally try to elect obviously incapable presidents?
Incumbent Presidents in the US almost always run for reelection to a second term, that's the normal expected outcome with vanishingly few exceptions in the last century.
And all the shenanigans around 2016 when Bernie Sanders was shut out by the DNC. Even Elizabeth Warren said outright that the DNC rigged the 2016 nomination. I think the outcome would have been very, very different if the DNC didn't fuck with the timeline and forced Hillary as the Dem nominee.
Sarbanes-Oxley? Dodd-Frank? ACA? Now I'm sure you (like, every other living human) has a ton of nitpicks and criticism of bills like that. But they're real and they passed and they're largely-to-wholely partisan creations of the Democratic policy apparatus (SOX looked very bipartisan and was passed under Bush, but the actual bill was written mostly by democratic staffers).
I mean, fine, you'd like different regulation. But they're responsible for basically all the corporate regulation active in the US regime.
Bothsidesism is a very poor tool for this particular argument. Work with your allies, please.
How’s ACA hard-hitting on corporations?
The 15/20 rules is what resulted in recorder profits in health insurance business, how’s this anti-corporate?
"You must sell insurance to sick people!", "I don't wanna!", "I don't care, the law now says you must."
I mean, the whole insurance industry had spent the better part of a century (basically its whole existence) denying coverage to people with pre-existing conditions, and the ACA fixed it via fiat.
Not only is it clearly "hard-hitting", I'd argue it's probably the single most "hard-hitting" invasive change to corporate regulation of the post-war era. Certainly it is in terms of the amount of money involved in the affected market.
> how’s this anti-corporate?
And this is where the rub is. You're not saying that your desired regulatory structure is something different than existing proposals, you're demanding that your government engage in behavior that is expressly "anti" whatever you perceive "corporate" interests to be.
That's not democracy, that's just warfare. You have enemies and you want to win. And the problem there is that the structure of our government requires compromise.
https://landline.media/affordable-care-act-tax-penalties-end...
>The penalty for not having an ACA-compliant plan or forgoing coverage in 2018 is the greater of $695 per adult and $347.50 per child for a maximum of $2,085 or 2.5 percent of adjusted household income. The percentage penalty is capped at an amount equal to the national average premium for a bronze plan, the minimum coverage available by law
If their profits are so good, their shares would be worth owning.
Could it be because their profit margins are actually terrible (3% if lucky), and they have to get the price of their product approved by 50 different state government employees, and the federal government constantly changes their customer’s subsidies making revenue and expenses extremely volatile and difficult to forecast?
If you're looking for policies to blame for insurance companies having a rough 2025-2026, look to the Republican-led efforts cap Medicare fees and alter eligibility requirements.
https://dqydj.com/sp-500-return-calculator/
https://dqydj.com/stock-return-calculator/
Since Jan 1, 2010, 18% for UNH, 15% for MOH/CNC, 14% for CI/ELV/HUM/SP500, 8.5% for CVS.
Also, these businesses are competing to sell a completely fungible commodity good with tons of non profits like Kaiser Permanente, HCSC, Cambia, Independence Blue Cross, and many others. It makes no sense to expect those businesses to be able to earn any decent profit margin when their competition only needs 0%.
And it is true, that it was signed by Clinton. But it was otherwise an entirely partisan bill passed with exclusively republican votes. All 44 democrats voted nay.
Now, given that, do you really feel that that's a supporting argument to the upthread contention that Democrats are a corporate-dominated party and not interested in regulation?
also if you look at how the two factions separate from each other decade by decade (there used to be overlap, almost none now) it's obvious that there will be no reconciliation.
it's a musical chairs game at this point, whoever is in power when the democratic facade cracks will be able to seize power and transition into whatever single party state follows.
They are not turning a profit, and aside from enterprise token billing, every other offering they have is a loss leader. In order to flip that around they need to drive down the cost of inference significantly, either through massive compute improvements, more efficient or lower cost models, and either one lowers the barrier to entry for local and/or private cloud inference for open models. The moat established by the expertise in model training is effectively dissolving as they approach recursive self improvement because any mechanism that prevents model distillation effectively reduces the quality and utility of the models, so each improved model makes their competitors better by default if they want to realize those gains.
The reason for the pivot to control and regulation seems to be more about erecting an artificial moat in their market, and having these technologies controlled or highlighted as supply chain risks will continue to drive investments in alternatives, especially in other countries where data sovereignty and domestic tech has become a priority given the last decade of the US dominated tech industry.
I don't think that AI is going to go away, but I think the future of AI in the next several years looks alot like massive overfunded foundational model providers collapsing, their employees launching dozens or hundreds of direct competitors, and the leading hyperscalers buying up datacenter capacity the way they bought up dark fiber 20 years ago.
Dems still subsidize O&G but they have other constituents they need to pay lip service to.
Banking and finance own both parties.
Isn't this basically what Anthropic wanted?
Which sounds sensible on the surface, until you realize that all military organizations have extensive experience with that exact question, and have spent centuries refining their practices. Deferring that to a private company is actually a major regression and a concerning role to pass on to the public regardless.
FWIW I am not pro military, but it's a prominent part of human culture and not going away anytime soon. I do understand the nature of professions, however, and only one of these groups has made the interpretation of this question their specialization.
I guess my opinion would really depend on the exact wording and details of the arrangement. Lots of people jumping in with strong stances based on pretty limited details.
This is false. Anthropic wanted assurance that its technology would not be used for fully autonomous weapons or domestic mass surveillance.
https://www.pearlcohen.com/anthropic-sues-department-of-defe...
It's not like Anthropic will be nationalized any more than the bankrupted auto companies of 2008.
It is impossible to miss the absolute hypocrisy of this moment and how people and corporations are supposedly free, but if you happen to be of interest to the government they will try to casually ruin you. It has always been this way, people just haven't noticed because it flies under the radar most of the time when the govern crushes some smaller enterprise or another because it got in their way or didn't play by their rules.
It seems simple to me. Anthropic is a private company. If you don't like it fix the laws or don't buy from them.
We've forgotten what it means to run a society, we're allowed to change rules how we see fit whenever we want to. Only the elites exercise this power, but people have it too.
In theory, maybe. In reality, this is a privilege for a very small number of rich people.
> Anything that erodes the rights of companies erodes your rights.
How absurd. Food safety regulations don't erode my rights one bit, but they do erode the rights of companies to negligently poison me. I guess I'm losing the right to start a food company that poisons people, but I'm happy to forfeit that right in exchange for not being poisoned. We could extend this list of reasonable tradeoffs in which we all exchange the right to treat people poorly for the right not be treated poorly, and we all would benefit.
Except the people who want to sell poisoned food, those people will be worse off. I am okay with that.
Anthropic’s position is also known as “having terms of service”.
To be clear the existing, signed ToS terms that Trump balked at were:
1. No mass surveillance of US civilians - illegal but we have all read the Snowden papers.
2. Fully unmanned kill chains on autonomous weapons - Anthropic have even stated they will support this eventually, they just didn’t think their models were ready and didn’t want the obvious, inevitable blowback.
The DoD is within its rights to cancel the contract if it doesn’t like the terms that it previously signed.
This seems like a definitional supply chain risk.
And it also doesn't mean that Anthropic is supply chain risk merely for not wanting AI to be the active decision maker in live or death situations in war.
And Anthropic's objection was recently confirmed: the US military blew up a school in Iran due to an AI error, an killed more than a hundred kids.
So yes obviously defense contractors are forbidden to use it.
The reasons why are incidental to that requirement for defense contractors.
What are you talking about? That's nowhere close to a true statement.
No, it does not mean that. I don't know why people have been making this claim. If you work for a company, let's call it X, and your work is not DOD related, let's say you work on a social media service, then X has no reason (from this designation) to prevent you from using Anthropic, it will not impact their defense contracts at all.
This designation means Athropic's systems can't be used in DOD work, but many companies that work with the DOD also do work outside that. That separate work is not restricted with respect to the use of Anthropic's systems. Repeating a lie as if it were true is just dumb, it helps spread that lie.
If you don't like that, and would prefer to discuss a lawful supply chain risk designation, you should convince him to retract the statements or impeach him.
Side note, the Anthropic bots are astroturfing the comments.
if this is more than "people i disagree with are commenting", you should be messaging hn@ycombinator.com rather than writing it as a side note in your comment.
by analogy it would be like the us military renting all m4 carbines from a political party.
anthropic treat customers very badly. they have claude lecture the user about the moral law according to anthropic. they change the product all the time. they have safety refusals to degrade functionality, they have mandatory logging so they spy on how it is used.
it's unusable for a serious application.
on top of that it is unsurprising that a big tyrant does not take well to a little tyrant.
it is a supply chain risk, so they don't want it in the supply chain.
that's logical.
they don't want suppliers to use it because it is a supply chain risk. specifically, the risk is that it is unreliable.
the designation stops suppliers from using it.
it stops it from being in the supply chain.
If Lockheed Martin disabled weapons systems when it discovered the Air Force using its fighters over Iranian airspace based on Lockheed's ideological views, I'm pretty sure they would be labeled a supply chain risk at minimum. What is the issue here?
fable 5.1 doesn't have any option to stop anthropic spying on it. you can't have that...anthropic contractors and actually the rationalist ea clowns themselves don't have security clearance
For all the downsides of losing the business of government contractors, it really would hold Anthropic to their stated values: they’ll never develop AI controlled weapons on this blacklist.
that's the mobster-loyalty mentality, and it's not how successful (read: desirable for the most people) governments function
it is a bizarre company which also inherits the authoritarian infrastructure of cloud services.
that is: remote control of the software, spying on customers, etc.
What the orange turd is going is rather petty bullshit in a market where many options exist.
But that's not at issue here. Obviously everyone agrees that the Pentagon can issue RFQ's to whatever specifications it wants. They can buy whatever they want, for the same reason that you or I are free to sell whatever we want.
The subject at hand is an attempt to use national security regulatory power to coerce Anthropic into selling a product it doesn't want to sell.
It would be naive to think that a private company could deny the government access to any tech if they want it for military or surveillance etc usage. In this case the government isn't forcing Anthropic to do it, they have alternatives and are perhaps making an example of Anthropic for refusing.
You could even make a case that the "supply chain risk" designation DOES make sense. If Anthropic are opposed to certain uses of their software, then doesn't that potentially impact companies that are part of the supply chain as well?
Is it coersion when people refuse to buy anything but free range chicken eggs?
DoD just refuses to buy a product. Anthropic just lost a customer because their product didn't fit the requirements of the customer.
Could you imagine a hypothetical where we're at war and a private company has a potentially war-ending tech (nuclear weapon say) that they refuse to sell to the government?
I suspect this will be reversed en banc.
Anthropic’s argument is really clear and easy to understand: we can’t simultaneously be a supply chain risk and also be a company the DoD demands we change our policies so they can use our models without restrictions.
but really, they all are, look what Grok did to hundreds of Iranian children
* https://thehill.com/policy/technology/5928204-pentagon-musk-...
Lol wtf, do you think nobody here knows how to use git? this is like a Fox News comment, you are not a developer and are trying to spread fear because a word was used jokingly. "git blame" is just a diff tool.
You cannot use git blame to cause conflict.