TechFlow news: On March 3, OpenAI CEO Sam Altman posted an internal announcement on social media stating that the company is revising its cooperation agreement with the U.S. Department of Defense (DoD) to explicitly prohibit the use of AI systems for domestic surveillance of U.S. citizens—including through commercial channels for acquiring personal information. The revised agreement also confirms that OpenAI’s services will not be used by DoD intelligence agencies (e.g., the NSA); any such use would require additional contractual amendments.
Altman acknowledged that issuing the announcement hastily on Friday was a mistake, noting that the issue is complex and requires clear communication. He emphasized that the company aims to act through democratic processes and stated that the technology is not yet mature—many safety concerns must be addressed incrementally with the DoD. Altman also expressed hope that competitor Anthropic would not be designated a Strategic Computing Resource (SCR) and would receive the same terms of cooperation.
