Fact-check: The article's claims about OpenAI's deal with the Pentagon for classified AI use, Sam Altman's 'definitely rushed' acknowledgment, stated protections against autonomous weapons and surveillance, and Anthropic concerns are corroborated by real-time X posts from March 4, 2026, including admissions of the rushed nature and backlash, as well as the source article itself and multiple web reports from MIT Technology Review, Economic Times, and Fortune. No contradictions found; consistent with current reporting amid US Military AI Policy developments.
The Download: The startup that says it can stop lightning, and inside OpenAI’s Pentagon deal
OpenAI has reached a deal with the Pentagon to allow the US military to use its technologies in classified settings, a move CEO Sam Altman acknowledged was 'definitely rushed.' The agreement includes stated protections against use for autonomous weapons and mass domestic surveillance, though it remains unclear whether safety precautions can be effectively implemented as the military pursues an AI strategy amid ongoing strikes on Iran. The deal also touches on concerns raised by Anthropic, which