OpenAI announces new deal with Pentagon — including ethical safeguards
Category: Policy
OpenAI struck a deal with the Pentagon on Friday night to deploy its models on classified networks — with written guarantees against mass domestic surveillance and autonomous weapons use. The catch that nobody can explain: these are essentially the same red lines that got Anthropic blacklisted hours earlier. Altman publicly called on the Defense Department to offer identical terms to all AI companies, a notable olive branch toward Anthropic, but the Pentagon hasn't explained why it accepted from OpenAI what it refused from its rival.