OpenAI's Altman defends Pentagon deal: 'You do not make operational decisions'
OpenAI CEO Sam Altman is standing by the company's new partnership with the Pentagon, making it clear that while OpenAI will provide AI expertise, it won't have a say in military operations.
As Altman put it, "You do not get to make operational decisions."
Deal was needed to ease tensions, says Altman
OpenAI agreed late Friday to let its AI models be used on classified Pentagon networks—a move that happened just after rival Anthropic turned down a similar deal and got flagged as a supply-chain risk by the Trump administration.
Altman admitted online that things felt "definitely rushed" but said the deal was needed to ease tensions.
Responding to critics, Altman pointed out several safeguards
The announcement sparked backlash over surveillance concerns, especially since Anthropic's Claude has now overtaken ChatGPT in Apple's App Store rankings.
Responding to critics, Altman argued this approach actually sets clearer boundaries than past deals.
He also pointed out safeguards: US laws against citizen surveillance and autonomous weapons still apply, technical safeguards are in place that OpenAI says will help prevent misuse, and cleared OpenAI engineers will work onsite alongside safety researchers.