OpenAI Details Pentagon AI Access
OpenAI has revealed more details about its landmark deal with the Pentagon, which grants the U.S. Department of Defense privileged access to its latest AI models. This strategic realignment follows the Trump administration's abrupt ban on rival AI firm Anthropic last week, showcasing how quickly the government is shifting its tech allegiances based on security concerns.
The dispute with Anthropic centered on the AI firm's refusal to remove safety restrictions from its Claude model. Anthropic insisted on prohibiting the use of its technology for mass domestic surveillance and in fully autonomous weapons, "red lines" the company argued were necessary because current AI is not reliable enough for such applications. In response, President Trump directed all federal agencies to cease using Anthropic's technology, and Defense Secretary Pete Hegseth designated the company a "supply-chain risk." This blacklisting, a measure historically reserved for foreign adversaries, gives agencies a six-month phase-out period and is being challenged in court by Anthropic. Hours after the ban, OpenAI announced its own agreement to deploy models on the Pentagon's classified networks. CEO Sam Altman stated the deal was "rushed" to "de-escalate" the conflict between the government and the AI industry. The agreement builds on a previous $200 million contract OpenAI signed in 2025 to develop AI for national security. OpenAI's deal notably includes contractual guardrails similar to those Anthropic requested, such as requiring "human responsibility for the use of force" and prohibiting mass domestic surveillance. The models will be cloud-based only, preventing their use on "edge devices" like drones where they could potentially operate as autonomous weapons.
