OpenAI Secures Major Deal with Pentagon as Trump, Hegseth Condemn Anthropic
OpenAI secures classified network access while Anthropic gets cut off for refusing surveillance and autonomous weapons terms.
In a significant shift in the U.S. government's AI strategy, OpenAI has secured a pivotal contract to deploy its advanced models on the Pentagon's classified networks. The agreement, which includes specific safety guardrails, allows defense teams to build applications for logistics, intelligence analysis, cybersecurity, and operational planning. This development comes immediately after the Trump administration moved to bar rival Anthropic from all federal use, citing national security concerns. OpenAI's deal builds on a previous $200 million Department of Defense contract from 2025 and positions the company as a primary AI provider within U.S. defense infrastructure, marking a dramatic consolidation of its government partnership.
The breakdown with Anthropic centered on two non-negotiable demands the company refused: the use of its AI for domestic surveillance and enabling fully autonomous warfare without human oversight. Anthropic CEO Dario Amodei stated these uses were incompatible with democratic values and that AI remains too unreliable for safe autonomous deployment. In response, President Trump directed federal agencies to cease using Anthropic's technology, and Defense Secretary Hegseth designated the company a Supply-Chain Risk to National Security, effectively blacklisting it from military contractors. Anthropic has condemned the designation as 'legally unsound' and plans to challenge it in court. This clash highlights the growing ethical and operational divide in how frontier AI models are integrated into national security, with OpenAI now positioned to shape that integration significantly.
- OpenAI secured a Pentagon deal for classified network AI deployment in logistics, intelligence, and cybersecurity.
- President Trump banned Anthropic from federal use, designating it a national security risk for refusing surveillance/autonomous weapons terms.
- Anthropic will challenge its 'Supply-Chain Risk' designation in court but must transition services within six months.
Why It Matters
This reshapes the defense AI landscape, setting precedents for ethical guardrails and determining which companies power national security.