OpenAI Revises Pentagon Deal After Internal and Public Backlash
OpenAI amended its Defense Department contract to bar intentional domestic surveillance after employee protests and app uninstalls, while the full contract remains unreleased and legal scrutiny continues.
CNBCSam Altman tells OpenAI staffers that military's 'operational decisions' are up to the government
TownhallOpenAI Adds Surveillance Ban in Deal With Pentagon
New York PostOpenAI’s Sam Altman fends off ‘painful’ backlash to Pentagon AI deal — including chalk-wielding activists
NBC NewsOpenAI alters deal with Pentagon as critics sound alarm over surveillance
Overview
Sam Altman told staff that OpenAI does not 'get to make operational decisions' about how the Department of Defense uses its AI, he said at an all‑hands meeting four days after OpenAI announced its DOD arrangement, according to a partial transcript.
OpenAI said it added language that 'the AI system shall not be intentionally used for domestic surveillance of U.S. persons and nationals' and that defense intelligence components are excluded, Sam Altman and Katrina Mulligan wrote on X.
More than 100 current OpenAI employees signed an open letter urging executives to refuse the Department of War's demands, and protesters surrounded OpenAI's San Francisco headquarters, sources said.
Last year OpenAI was awarded a $200 million contract by the Pentagon, and Sensor Tower data showed the ChatGPT app's daily uninstall rate rose by about 200% after the announcement.
The full contract text has not been released, and legal experts said greater transparency is needed to evaluate the safeguards and potential carve‑outs in the agreement.
Analysis
Center-leaning sources frame the story as a tech-industry defense partnership controversy, emphasizing procedural reassurance and competitive context. Editorial choices—timing cues, prioritizing Altman's conciliatory remarks and unnamed insider details, and highlighting employee criticism plus rival firms like Anthropic—tilt coverage toward operational and market implications rather than deep ethical or policy critique.
FAQ
The contract is to develop prototype frontier AI capabilities, including agentic AI workflows, for national security challenges in warfighting and enterprise domains, with completion expected by July 2026.
OpenAI added language barring intentional use of the AI system for domestic surveillance of U.S. persons and nationals, and excluded defense intelligence components; deployment is cloud-only with OpenAI safety experts involved.
Over 100 employees signed an open letter protesting, protesters surrounded OpenAI's headquarters, and ChatGPT app daily uninstalls rose by about 200% after the announcement.
No, the full contract text has not been released, and legal experts call for greater transparency to evaluate safeguards and potential carve-outs.
