Anthropic co-founder and CEO Dario Amodei isn’t happy with OpenAI head Sam Altman (perhaps predictably). In a memo to staff reported by The Information, Amodei called OpenAI’s deal with the Department of Defense “safety theater.”
“The main reason[OpenAI]accepted[the Department of Defense agreement]and we didn’t is because they cared about appeasing their employees, and we actually cared about preventing abuse,” Amodei wrote.
Last week, Anthropic and the US Department of Defense (DoD) failed to reach an agreement over the military’s request for unrestricted access to the AI company’s technology. Anthropic already has a $200 million contract with the military, but the Pentagon has vowed it will not use its AI for domestic mass surveillance or autonomous weapons.
Instead, the Department of Defense (known as the Department of the Army under the Trump administration) contracted with OpenAI. Altman said his company’s new defense contract would include protection against the same red lines that Anthropic insisted on.
In a letter to staff, Amodei called OpenAI’s message a “blatant lie” and falsely claimed that Altman “is presenting himself as a peace negotiator and deal negotiator.”
Amodei may not be simply speaking from a place of bitterness here. Anthropic specifically took issue with the Department of Defense’s claim that its AI could be used for “any lawful use.” OpenAI said in a blog post that the contract allows it to use its AI systems for “all lawful purposes.”
“In our communications, it was clear that the DoW considers large-scale domestic surveillance to be illegal and has no plans to use it for this purpose,” OpenAI’s blog post said. “We made sure that the contract spelled out the fact that it was not subject to legal use.”
tech crunch event
San Francisco, California
|
October 13-15, 2026
Critics say the law is subject to change, and what is currently illegal could be allowed in the future.
And the public seems to be on Anthropic’s side. After OpenAI was awarded a contract with the Department of Defense, ChatGPT uninstalls spiked by 295%.
“I don’t think this spin/gaslighting attempt is playing very well in the general public or the media, where people primarily see OpenAI’s dealings with DoW as opaque or questionable, and see us as heroes (we’re currently #2 on the App Store!),” Amodei wrote to staff. “It’s working for some Twitter idiots, which is fine, but what worries me most is how do I make it not work for OpenAI employees?”
