[flagged]
Except it's not, really. If Anthropic/Claude doesn't mean the DoD's need, they can and should just put out an RFP for other LLM providers. I'm sure there's plenty of others that'd happily forgo their morals for that sweet government contract money.
No US company has to provide services to the DoD or any other branch of government. It's not "veto power" it's being selective of who you do business with, which is 100% legal.
> has never been invoked against an American company.
There's always a first. I am assuming it is not illegal to do that. It's a completely reasonable business decision to ensure your supply chain does not depend on things that may change against your goals. For example, you don't want to build or depend on an open source platform that you know is gonna rug pull, if you count on it remaining open source, do you? American or otherwise.
I don't understand this phenomena of people acting like the stupidest human beings who ever lived. This is not your first day on Earth - you understand how coersion works and what voluntary means.
As to this being 100% legal, I'm not so sure (not a lawyer). It might not be a criminal offese, but there's a whole category of abuse of power that this may fall under if Anthropic is put under a certain status without real justification. Many powers given to the executive branch are not absolute and can't be applied arbitrarily, but require justification. Anthropic might be able to sue the government for declaring them a "supply-chain risk" without sufficient justification. E.g. they could claim that not being sufficiently patriotic in the eyes of the administration does not constitute a risk, and that since their not the sole supplier of the tech, they were not trying to strong arm the government to do anything.
> does not "influence their supply chain"
I would be wary of making this conclusion. Obviously it could conceivably influence the supply chain when you build on top of their model. If you look at the type of risks enumerated in DoD guidelines, it is not just "oh this software has vulnerability" which is what started the discussion in this subthread in the first place. There are many kinds of risks DoD needs to address, none are particularly new; including Sustainment Risk. The closest thing I remember to this case was Sun Java "no use in nuclear facility" EULA term, which LLM suggests was ignored by DoE/D because that was interpreted as a "limitation on warranty" not a "restriction of use."
I understand 'goals' and 'means to an end', but this concept of "law" evades me.
> Anthropic's terms were laid out in the contract the Pentagon signed, which they want to forcibly amend.
It's called negotiation in business. I am sure both sides are clear-eyed on what the consequences were and Anthropic made a calculated bet (probably correctly) that some segment of their employee/customer base would get wet by hearing this news and it more than offsets the lots business, thus is worth it.
Companies have gone out of business due to a big customer pulling the contract. Imagination Technologies comes to mind. This is not a rare thing in business.
> full power of the US government
Haha, I can assure you that is not even close to the full power of US government. Ask the crypto people during Biden admin for just a little more power (still not even close to "full.")
For a company of Anthropic's size, this may very well be a death sentence, even if their work has nothing to do with the military supply chain. They could have just canceled the contract, but they wanted to go full Darth Vader on them to prove a point in case anyone else thought about "negotiating" "voluntarily" with the federal government.
People have noticed.
> It's called negotiation in business.
The bad faith in this statement alone is almost equal to the sum of it in the rest of your comments.
The contract, including Anthropic's redlines, was signed more than a year ago and has been humming along with no objections from anybody. Hegseth abruptly got a bug up his ass about it last week, and demanded Anthropic sign a revised version under threat of punishment. Anthropic is simply saying "no, we will not be forced into signing a new version, you can either keep going with the original terms we all agreed to, or stop using us". The Pentagon can simply stop using Anthropic if they don't like the terms anymore (which, again, are the terms Pentagon agreed to in the first place). But what the DoW wants is to strong-arm Anthropic, using the DPA, into new terms because they abruptly changed their mind. That's not "negotiation" in any sense, that's Mafia behavior.
The actual terms of the contract aren't even relevant, this is purely a matter of tort law and whether you can bully someone into a new contact because you woke up one day and decided you didn't like the one you agreed to.
> Anthropic's terms were laid out in the contract the Pentagon signed, which they want to forcibly amend.
They want to "forcibly amend" is either within their rights per original contract, or not. One is fair game, the other is not.
Isn't agreeing to amend a contract always within their rights?
That said, many government contracts include some variant of "we can cancel at any time for any reason".
I’d agree it is a serious risk.
The current government is deeply unpopular, it's only going to get worse for them.