OpenAI Rewrites 'Sloppy' Pentagon AI Deal After Backlash Over Surveillance Risks

OpenAI is revising its contract with the US Department of War following public criticism and employee concerns about domestic surveillance. The initial agreement, replacing Anthropic, was rushed and perceived as opportunistic. CEO Sam Altman stated the revised contract would explicitly forbid the intentional use of AI systems for domestic surveillance of U.S. citizens. Intelligence agencies like the NSA are excluded from the deal's initial scope. The Department of War and OpenAI are adding explicit language to reaffirm the commitment for their AI tools to not be used on U.S. persons. The Department of War is convening a working group for ongoing dialogue on AI, privacy, and security challenges, with OpenAI participating. OpenAI plans to rely on contractual provisions and technical controls to uphold its red lines. The controversy arose after Anthropic's negotiations with the Pentagon collapsed over similar concerns regarding mass surveillance. The fallout includes government actions against Anthropic and internal dissent within OpenAI. Employees at OpenAI and Google signed an open letter, expressing concern over government demands. Public protests and consumer shifts have also emerged as a result. Miles Brundage, OpenAI's former head of policy research, criticized the company's handling of the negotiations.