OpenAI and Palantir Win Pentagon Deals Amid Anthropic Ban
In the wake of the White House ordering agencies to cut ties with Anthropic, OpenAI has secured a significant Pentagon AI contract. Meanwhile, the DoD awarded Palantir a $7.6 million contract for data integration, bypassing the competitive tender process due to urgent need. The moves underscore how quickly the Pentagon is shifting its AI supply chain based on compliance and perceived security risks.
The directive to blacklist Anthropic originated from a standoff over the company's refusal to permit its AI models to be used for mass domestic surveillance or fully autonomous weapons. This culminated in the White House designating the AI firm a "supply chain risk," an unprecedented action historically reserved for foreign adversaries, not American tech companies. There is a six-month phase-out for the Department of Defense to transition away from Anthropic's technology. OpenAI’s resulting contract to work with the Pentagon on its classified network incorporates the very "red lines" Anthropic advocated for. The agreement explicitly prohibits the use of OpenAI's models for mass domestic surveillance, to direct autonomous weapons, or for high-stakes automated decisions. Unlike the proposed terms with Anthropic, OpenAI's deployment is restricted to the cloud, with cleared company personnel remaining in the loop and retaining control over the safety stack. Palantir's recent sole-source contracts with the DoD are often justified by "urgent and compelling need," allowing the government to bypass standard competitive bidding. This justification is typically used when a vendor's proprietary technology, like Palantir's Gotham platform for data integration and analysis, is deemed irreplaceable for a critical mission, or when waiting for a competitive process would cause unacceptable delays. These contract maneuvers align with the Pentagon's aggressive "AI-First" strategy, which prioritizes rapid experimentation and fielding of commercial AI capabilities to maintain a competitive military edge. A key mandate of this new doctrine is to make the latest AI models available to military users within 30 days of their public release, establishing deployment speed as a primary procurement criterion.
