OpenAI Promises AI Won't Spy on You While Secretly Signing Pentagon Deal

KEY POINTS

•On March 3, 2026, OpenAI CEO Sam Altman announced contract changes with the Pentagon after public backlash over surveillance fears.
•Altman emphasized their AI won't be used for domestic surveillance or by Department of War intelligence, citing laws like the Fourth Amendment.
•Protests occurred in San Francisco and London alongside employee support for rival Anthropic's stricter AI military usage bans.

On March 3, 2026, OpenAI's CEO Sam Altman performed the corporate version of a facepalm, announcing amendments to their freshly inked Pentagon contract after a public freakout over AI mass surveillance fears. After debuting the deal on a Friday — conspicuously just a day before the US bombed Iran — Altman admitted they 'rushed' into an agreement that, by Monday evening, needed quick fixes to reassure jittery Americans their AI won't be used against them domestically or by war intelligence groups like the NSA (unless another contract magically appears). Meanwhile, protests exploded in London and San Francisco, with nearly 500 OpenAI and Google employees siding with Anthropic’s red lines, showing the moral grand prix in AI warfare just got haphazardly weird.