OpenAI's Altman Seeks Middle Ground in Defense AI Standoff, Backs Anthropic's Stance

In a move to calm rising industry tensions, OpenAI CEO Sam Altman told employees Thursday that his company should attempt to ease a standoff between rival Anthropic and the Department of Defense. The statement came as a Friday deadline loomed for Anthropic to decide on a Pentagon request for unrestricted, lawful use of its AI models.

Anthropic has sought binding assurances that its technology will not be deployed for autonomous weapons or mass domestic surveillance, conditions the DOD has so far not accepted. In an internal memo viewed by CNBC, Altman outlined OpenAI's own 'red lines,' writing, 'We have long believed that AI should not be used for mass surveillance or autonomous lethal weapons, and that humans should remain in the loop for high-stakes automated decisions.'

Before Altman's memo, approximately 70 OpenAI staffers had signed an open letter supporting Anthropic, titled 'We Will Not Be Divided.' Altman, in a Friday interview, expressed trust in Anthropic's safety commitments and support for military personnel, but admitted uncertainty about the outcome.

The situation highlights a complex moment for AI firms with government ties. OpenAI holds a $200 million DOD contract for nonclassified uses, while Anthropic was the first to integrate its models into classified mission workflows. Altman stated OpenAI is exploring a potential agreement to operate in classified environments with technical safeguards and personnel oversight, explicitly excluding domestic surveillance and autonomous offensive weapons.

'This is a case where it's important to me that we do the right thing, not the easy thing that looks strong but is disingenuous,' Altman wrote, acknowledging the short-term optics might be challenging. Further discussions with internal safety teams were scheduled for Friday.

Source: CNBC