Tommy Vitor

data of some sort, but who knows?

I'm literally guessing.

Regardless, though, those news reports or maybe some conversation between an Anthropic executive and some Palantir executive kicked up this broader conversation between Anthropic and the Pentagon about the use of Claude and when it violated their terms of service.

For context, despite having a

that seems to be the most concerned with safety and how technology is used and are the most openly in favor of government regulation, although you never know with these guys.

They can change their mind.

The Pentagon has contracts with the four major AI companies, OpenAI, XAI, Google, and Anthropic.

Apparently, Hexeth's team went to all of them and said, basically, we want to rewrite your terms of service and replace them with language that says the U.S.

military can do anything lawful with your models.

OpenAI, XAI, Google, they were like, fine.

Anthropic said, no, we want to make sure our tech is not used for mass surveillance of Americans or in autonomous killing machines, like killer robots.

And the Pentagon basically replied, how fucking dare you?

Kind of a tell.

Yeah.

And so the Pentagon threatened not only to cancel Anthropix, I think it's like $200 million contract with the Pentagon, but also to designate them as a supply chain risk, which is a step in the past the US has done to Huawei, which is a Chinese telecom company.

and Kaspersky Lab, which is a Russian antivirus company.

So big, like, state-backed companies that we were worried could be backdoors for state actors.

That, we don't have to get into all the details, but that could create huge business problems for Anthropic.

On Tuesday, the day we recorded this, Anthropic CEO Dario Amadei met with Pistol Pete, Secretary Hegseth,

According to a readout they leaked to CNN, Hegseth gave Anthropic until Friday to get rid of its safeguards or else he will punish them, as I mentioned earlier.