OpenAI Inks Deal for Classified Military AI
OpenAI has reached an agreement to deploy its AI models on classified Department of Defense networks. The deal marks a major step in integrating commercial generative AI into sensitive national security operations and reportedly includes strict "guardrails" for its use.
The Pentagon's push for artificial intelligence is part of a broader "AI Acceleration Strategy" mandated by President Donald Trump to create an "AI-first" warfighting force. This strategy aims to speed up experimentation with advanced AI, eliminate bureaucratic hurdles, and increase investment to maintain a military advantage. OpenAI's agreement is valued at up to $200 million and is with the Pentagon's Chief Digital and Artificial Intelligence Office. The company will develop "prototype frontier AI" for both warfighting and internal administrative functions, with the work expected to be completed by July 2026. The deal followed the breakdown of negotiations with rival AI company Anthropic. The Pentagon had demanded that Anthropic remove its safeguards against the use of its AI for mass domestic surveillance and fully autonomous weapons, a request Anthropic refused, stating it could not "in good conscience" agree. Consequently, the Trump administration ordered federal agencies to stop using Anthropic's technology. In response to public backlash after securing the deal, OpenAI CEO Sam Altman admitted the announcement looked "opportunistic and sloppy." The company then amended the agreement to include specific "red lines," explicitly prohibiting the use of its technology for mass domestic surveillance, fully autonomous weapons, and high-stakes automated decisions. As part of the agreement, OpenAI will deploy its models via the cloud, retaining control over the safety features, and will have its own engineers on-site to oversee compliance. The deal also explicitly bars U.S. intelligence agencies, such as the NSA, from using OpenAI's services under this contract. This is not the first time a major tech company has navigated the ethics of military AI work. In 2018, Google withdrew from the Pentagon's Project Maven, an initiative to use AI to analyze drone footage, after thousands of its employees protested, stating that "Google should not be in the business of war."
