Sam Altman told OpenAI staff he was trying to "save" Anthropic as the Pentagon severed ties with the rival AI company, internal Slack messages reveal. What followed was a masterclass in corporate positioning: Altman portrayed himself as a stabilizing force even as OpenAI swooped in to claim the defense contract Anthropic had lost.
The messages, reviewed by Axios, show Altman's thinking shifted between February 24 and March 2 as a public dispute between the Pentagon and Anthropic escalated into one of the tech industry's most visible collisions with federal power.
On Feb. 24, as anxiety rippled through OpenAI's ranks, Altman suggested the moment was ripe to negotiate a classified systems contract with Defense Under Secretary Emil Michael. Michael called Altman that afternoon, and the two sides exchanged draft terms the next day.
By Feb. 26, Altman addressed the full staff with a careful message: OpenAI shared Anthropic's concerns about the Pentagon's demands and wanted to help defuse the situation. He also made clear that OpenAI intended to strike its own deal. Altman acknowledged the appearance might look uncomfortable short-term but framed his approach as principled rather than opportunistic.
The narrative shifted again on Feb. 27. Altman told a smaller group that Pentagon-Anthropic talks had deteriorated, suggesting CEO Dario Amodei was
Comments