After the Pentagon, OpenAI seeks to strike a deal with NATO—Altman urged staff to refrain from discussing politics

OpenAI and the U.S. military: a recap

What happened Key details Problem In a recent deal between OpenAI and the U.S. Department of Defense protests arose among startup employees. Leadership Decision CEO Sam Altman asked the team not to comment on Pentagon actions and emphasized that operational decisions would remain with American authorities. Company plans Rumors say OpenAI is trying to negotiate with NATO for supplying its AI solutions in non‑classified systems.

1. Employee meeting
Yesterday, according to CNBC, an all‑staff meeting was held at OpenAI. The recent deal with the Pentagon was discussed. Altman repeated last week’s stance – the company does not get involved in how the military uses its technology. He gave an example: “You can consider a strike on Iran correct and an invasion of Venezuela wrong, but that’s for you to decide.”

2. Pentagon role
The Pentagon acknowledged OpenAI’s technical expertise and will follow the company’s recommendations when optimizing AI technologies. Final decisions on their use remain with the head of the military – Pete Hegseth. Since last year the Pentagon has used OpenAI solutions in open projects, and a $200 million contract now allows AI to be applied in secret operations.

3. Restrictions and goals
Altman stressed that OpenAI aims to prevent its systems from being used for mass surveillance of U.S. citizens and automated target selection by weapons. The company wants to limit military use of AI to the department’s cloud infrastructure, not end devices.

4. Future partners
During the meeting Altman mentioned a potential contract with NATO. This is logical given the role of the American military complex in the alliance. The Pentagon considers AI suitable for any lawful purpose, but OpenAI and competitors such as Anthropic demand strict limits.

5. Summary
The agreement with the Department of Defense was “difficult,” but Altman sees it as a right step, even if it may temporarily affect OpenAI’s image. He believes the company can offer the government better AI models and possibly bring new players into cooperation, such as xAI, which are ready to meet any requests.

In short:

OpenAI agreed with the Pentagon to use its technologies in secret projects but left operational control to U.S. authorities. The CEO urged employees not to comment on Defense Department actions and announced plans to sign a deal with NATO. The company seeks to restrict military AI use to cloud services and avoid automated weapons.