OpenAI20:38PolicyOfficial Blog
OpenAI-DoW Safety Agreement for Classified AI
Maintains safety standards even in military use, boosting model trust.
Key Points
- 1Bans surveillance, autonomous weapons, high-risk decisions
- 2Cloud-only with OpenAI safety control
- 3Terms available to all AI firms
- 4Multi-layer protection via laws/contracts
OpenAI announced an agreement with DoW for deploying advanced AI in classified settings. It bans mass surveillance, autonomous weapons, and high-stakes automation. Cloud deployment keeps OpenAI in control of safety, with no impact on general users.