OpenAI Deploys AI Models in Pentagon Networks
In a significant development, OpenAI is deploying its AI models within the Pentagon's classified networks. The deal reportedly includes human oversight limits on surveillance applications. This move comes amid ongoing ethical debates and is seen as a driver for some government entities shifting to competitors like Anthropic's Claude for sensitive workloads.
The deal materialized after the Pentagon's negotiations with competitor Anthropic collapsed. Anthropic had sought to contractually prohibit its Claude AI from being used for mass surveillance or in autonomous weapons systems without human input, leading the Trump administration to order federal agencies to stop using its technology and designate it a "supply chain risk." OpenAI CEO Sam Altman stated the agreement includes specific "red lines" and more safeguards than previous military AI contracts. The contract prohibits the use of OpenAI technology for mass domestic surveillance, to direct autonomous weapons, or to make high-stakes automated decisions, with the company retaining control over its safety systems and keeping cleared personnel involved. The AI models will operate on a secure, private network to assist with office and planning work, such as managing large datasets, writing reports, and translating languages. This function supports long-standing initiatives like Project Maven, a program founded in 2017 to use machine learning to analyze surveillance imagery and flag potential targets for human
