Leaving autonomous weapons aside, how does Anthropic justifies that they signed up with surveillance company Palantir and now raising concerns for same surveillance with DoD?
This is very easy to explain. Anthropic outlines some limitations in their terms of service. Palantir accepted those terms. The DoD did not.
OpenAI claims their terms of service for DoD contain the same limitations as Anthropics proposed service agreement. Anthropic claims that this is untrue.
Now given that (a) the DoD terminated their deal with Anthropic, (b) stated that they terminated because Anthropic refused modify their terms of service, and (c) then signed a deal with openAI; I am inclined to believe that there is in fact a substantial difference between the terms of service offered by Anthropic and OpenAI.
Yeah, it never made sense when Sam immediately said that they had the same constraints yet de DoW immediately agreed with that.
From what I can see, OpenAI’s terms basically say “need to comply with the law”, which provides them with plenty of wiggle room with executive orders and whatnot.
Are you sure about that? Every information I’ve seen suggests that the DoD has been using Anthropic’s models through Palantir.
My understanding is that Anthropic requested visibility and a say into how their models were being used for classified tasks, while the DoD wanted to expand the scope of those tasks into areas that Anthropic found objectionable. Both of those proposals were unacceptable for the other side.
Wasn’t the trigger for all this what happened with Maduro earlier this year? From what I understood, Anthropic wasn’t very happy how their systems were being used by the DoW through Palentir which caused this whole feud.
And why would they have an objection to that? They sold a product to a customer. They should have no business in how that customer uses their software.
> And why would they have an objection to that? They sold a product to a customer. They should have no business in how that customer uses their software.
They sold a service to a customer, contractually subject to terms they both agreed upon. How do people keep missing this? The government changed their mind after agreeing to the restrictions and tried to alter the deal with Anthropic ex-post-facto.
It’s a bit more complex than that, but to be fair I don’t know what they were expecting after they integrated a purpose-built model with Palantir to be deployed in high-security networks to carry out classified tasks.
I'd hate to break it to you, but companies do have a right to determine how their products are used. You were subject to that when you wrote that comment. Did you not notice that?
“We’ve actually held our red lines with integrity rather than colluding with them to produce ‘safety theater’ for the benefit of employees (which, I absolutely swear to you, is what literally everyone at [the Pentagon], Palantir, our political consultants, etc, assumed was the problem we were trying to solve),” Amodei reportedly wrote.
“The real reasons [the Pentagon] and the Trump admin do not like us is that we haven’t donated to Trump (while OpenAI/Greg have donated a lot),” he wrote, referring to Greg Brockman, OpenAI’s president, who gave a Pac supporting Trump $25m in conjunction with his wife.
Another reason is that Sam Altman has been willing to "play ball" like providing high-profile (though meaningless) big announcements Trump likes to tout as successes. For example:
> "The Stargate AI data center project worth $500 billion, announced by US President Donald Trump in January 2025, is reportedly running into serious trouble.
More than a year after the announcement, the joint venture between OpenAI, Oracle, and Softbank hasn't hired any staff and isn't actively developing any data centers, The Information reports, citing three people involved in the "shelved idea."
Reminds me of when they cut the camera to Zuck and he made the $600 Billion Deal announcement, but was hot mic'd after and said "I'm sorry I wasn't ready... I wasn't sure what number you wanted to go with". I will be extremely surprised if half of these deals actually go through
Just to nitpick, Palantir isn't doing surveillance like Flock. They do data integration the way IBM does under contract for the governments. Some data pipelines include law enforcement surveillance data which get integrated with other software/databases to help police analyze it. There's no evidence they are collecting it themselves despite recent headlines. It's a relatively minor but important distinction IMO.
It’s the same with Facebook selling user data. Neither selling your data, like the carriers do, or selling the ability to target you with your data, like Facebook does, are very nice. But legally they are separate things that need to be regulated differently. As is the case with Flock and Palantir.
The solution is still no different than a decade ago. Far stricter laws on intelligence, federal and local police surveillance, and a reduction in executive power which oversteps checks and balances.
There will always be another IT company willing to do integrations even if Palantir dies. Software isn’t going away.
Right. But this is about Anthropic -- a company frames itself as a responsible and ethical steward of LLM technology. They can't pretend that OpenAI is somehow morally bankrupt here while continuing to deal with companies that undermine peoples' civil liberties.
I'm also a little unsure what you're saying here. Are you saying that it's futile to rely on corporate leaders to commit to ethical acts, as there's always someone else who will debase themselves to make money? I think that solely relying on the state to regulate itself with respect to civil liberties is a fast path to despotism. The well-regulated state was always a partnership between ordinary people bravely standing up for their rights and the norms of the rules and laws that made it socially acceptable to do so.
If I'm grasping you correctly, I think you're right; however, this points to the rottenness of our culture's way of organizing labor: the optimization of the shareholder over everyone else leads to some really awful effects.
I think a company which provides a sensor fusion dragnet for a government-run mass domestic civilian surveillance system is at least as culpable (and odious) than the ones supplying the data.
Though, I guess IBM did get away with lots of stuff that... Actually, did any supply companies in the WWII German war machine actually get in trouble for war crimes, or did they just go after officers and the people actually working in the camps?
The company selling punchcards that were used for logistics was apparently fine. What about the people making the gas canisters, or supplying plumbing fixtures? The plumbers? Where's the line?
Wondering, since this is increasingly becoming a current events question instead of an academic concern.
There were the so-called Subsequent Nuremberg Trials (12 of them). Among them were the trials of IG Farben (gas chamber supplies, Zyklon B) and Krupp (armament of the German military forces in preparation of an aggressive war)
I'm under no illusion that all the perpetrators of war crimes were held accountable but it's not a bad model.
Sure, but it's not as if the DoD was planning on using Anthropic to _collect_ the data either? I assume that the hypothetical DoD use case Anthropic shied away from dealt with the processing of surveillance data, just like what Palantir does.
> The military’s Maven Smart System, which is built by data mining company Palantir, is generating insights from an astonishing amount of classified data from satellites, surveillance and other intelligence, helping provide real-time targeting and target prioritization to military operations in Iran, according to three people familiar with the system...
> As planning for a potential strike in Iran was underway, Maven, powered by Claude, suggested hundreds of targets, issued precise location coordinates, and prioritized those targets according to importance, said two of the people.
It might match. The red line was domestic surveillance. You don't know what deal they had. Giving Anthropic the benefit of the doubt, perhaps Palantir said "Deal, we won't use your tool domestically".
> Who in their right mind would give the benefit of the doubt?
I'm saying that we should give Anthropic the benefit of the doubt that when they say "our deal with Palantir doesn't cross our red line", we should believe Anthropic, that they have gotten an assurance from Palantir that they wouldn't use it domestically. I'm NOT saying we should give Palantir the benefit of the doubt.
I wasn't commenting on "is giving AI to Palantir a good idea" (I don't think it is), I was commenting on "should we conclude that Anthropic is being dishonest because they claimed they have red lines but work with Palantir" (I think it's unclear, but there's a plausible explanation in which they're not being dishonest, but possibly naive, so give them the benefit of the doubt).
Whether you disagree with whether it truly aligns with their stated values, in their partnership with Palantir (making Claude available within their AI platform) they requested consistent restrictions:
> “[We will] tailor use restrictions to the mission and legal authorities of a government entity” based on factors such as “the extent of the agency’s willingness to engage in ongoing dialogue,” Anthropic says in its terms. The terms, it notes, do not apply to AI systems it considers to “substantially increase the risk of catastrophic misuse,” show “low-level autonomous capabilities,” or that can be used for disinformation campaigns, the design or deployment of weapons, censorship, domestic surveillance, and malicious cyber operations.
The moral disposition of the Anthropic leaders doesn't matter because they don't own the company. Investors won't idly watch them decimate billions in ROI by alienating the largest institutional customers on the planet.
> The moral disposition of the Anthropic leaders doesn't matter because they don't own the company. Investors won't idly watch them decimate billions in ROI by alienating the largest institutional customers on the planet.
Anthropic is a Public Benefit Corporation chartered in Delaware, with an expressed commitment to "the responsible development and maintenance of advanced AI for the long-term benefit of humanity."
So in theory (IANAL), investors can't easily bully Anthropic into abandoning their mission statement unless they can convince a court that Anthropic deliberately aimed to prioritize the cause over profit.
> So in theory (IANAL), investors can't easily bully Anthropic into abandoning their mission statement unless they can convince a court that Anthropic deliberately aimed to prioritize the cause over profit.
So why were they ever working with the military in the first instance, if that's the case? If you didn't gleam from OpenAI that it doesn't matter. Everyone is greedy and will jump ship for money if Anthropic does not get it for them.
This exchange between Anthropic and OpenAI feels a lot like theater. If I was really trying to stop abuses I wouldn't going out of my way to talk about it. The "public sees us as the hero's" bullshit feels like a smoke screen. Id make one statement and keep silent and let the public do the math and not get involved.
OpenAI claims their terms of service for DoD contain the same limitations as Anthropics proposed service agreement. Anthropic claims that this is untrue.
Now given that (a) the DoD terminated their deal with Anthropic, (b) stated that they terminated because Anthropic refused modify their terms of service, and (c) then signed a deal with openAI; I am inclined to believe that there is in fact a substantial difference between the terms of service offered by Anthropic and OpenAI.