OpenAI’s Sam Altman told staffers that he wants to “help de-escalate” the spat between the Pentagon and its chief AI rival Anthropic – which faces being blacklisted after saying it won’t remove military safeguards ahead of a key Friday deadline.
Altman told staffers in a memo late Thursday that OpenAI shares the same safety “red lines” as Anthropic – namely, that its AI tools should not be used for mass surveillance of Americans or to power weapons capable of firing without human oversight.
OpenAI is in talks with the Pentagon on a potential deal to allow use of its AI models in classified military settings, according to Altman – who wrote that the discussions were important to reach a consensus and avoid setting a dangerous industrywide precedent.
“Regardless of how we got here, this is no longer just an issue between Anthropic and the [Pentagon]; this is an issue for the whole industry and it is important to clarify our stance,” Altman wrote in the memo, according to Axios.
At the same time, Altman suggested that OpenAI could reach common ground with the Pentagon on how models should be used without violating its red lines.
“We believe this dispute isn’t about how AI will be used, but about control. We believe that a private US company cannot be more powerful than the democratically-elected US government, although companies can have lots of input and influence. Democracy is messy, but we are committed to it,” Altman wrote.
The Wall Street Journal was first to report on the memo.
Anthropic, whose Claude chatbot is the only model currently used by the US military in classified settings, has until 5 pm ET on Friday to remove safeguards. Defense Sec. Pete Hegseth said he will either classify Anthropic as a “supply chain risk” or invoke the Defense Production Act to force it to tailor models for miliary use unless it complies.
Altman’s memo surfaced hours after Anthropic CEO Dario Amodei that his safety-obsessed firm “cannot in good conscience accede” to the Pentagon’s demands.
Amodei said the Pentagon’s latest offer in tense negotiations “made virtually no progress on preventing Claude’s use for mass surveillance of Americans or in fully autonomous weapons.”
Emil Michael, defense undersecretary for research and engineering, blasted Amodei over the stance, writing on X that the Anthropic boss “has a God-complex” and “wants nothing more than to try to personally control the US Military and is ok putting our nation’s safety at risk.”
OpenAI’s potential intervention comes amid tense relations between Altman and Amodei – who recently refused to shake hands on stage during a viral photo op at an AI summit in India.
Earlier this week, a senior Defense official told The Post that Elon Musk’s Grok has received approval for use in classified settings, while OpenAI and Google were “close.”
