## Pentagon Clears Seven AI Vendors for Classified Work — Anthropic Excluded as Supply-Chain Risk
The U.S. Defense Department has finalized AI agreements with seven technology providers — OpenAI, Google, Microsoft, Amazon, Nvidia, Elon Musk's xAI, and the startup Reflection — granting the Pentagon access to their AI systems for classified operations, according to an announcement on Friday. The deals mark a significant expansion of the military's commercial AI partnerships, building on earlier agreements with OpenAI and xAI. Google has separately struck a similar arrangement, as reported by The Information.

The most notable development is the exclusion of Anthropic. The company, which previously worked with the Defense Department on classified information tasks, has been declared a supply-chain risk — effectively cutting off a former ally from the Pentagon's classified AI ecosystem. The move raises questions about what specific vulnerabilities or dependencies triggered the risk designation, and how Anthropic's technical approach diverges from the cleared vendors in ways that prompted concern.

The vendor selections signal a strategic narrowing of the Defense Department's AI supply chain. The Pentagon appears to be consolidating relationships with providers whose development practices and infrastructure meet its threshold for secure, controlled deployment in sensitive environments. xAI, which had already signed on for lawful use of its systems, secured a place alongside more established players — while Anthropic, despite demonstrated classified capabilities, finds itself outside the emerging framework. The composition of cleared vendors spans hyperscalers, chip manufacturers, and smaller AI startups, suggesting the Pentagon is building redundancy across the supply chain while maintaining strict controls over which systems operate in classified settings.

The implications extend to the broader defense AI landscape. With major vendors now formally aligned with Pentagon requirements, the cleared vendor list could become a de facto standard for which AI companies are trusted for national security applications — raising pressure on firms outside the circle to address supply-chain concerns or risk being locked out of significant government work.
---
- **Source**: The Verge
- **Sector**: The Network
- **Tags**: defense AI, Anthropic, OpenAI, Google, Pentagon contracts
- **Credibility**: unverified
- **Published**: 2026-05-01 14:54:08
- **ID**: 78864
- **URL**: https://whisperx.ai/en/intel/78864