Observed Signal · Mar 4, 2026 · Policy Update · Source: CNBC Technology · Impact: 4/5 · Sentiment: Negative
Defense Firms Abandon Claude After Pentagon Blacklist
The Trump administration announced a supply-chain risk designation for Anthropic, prompting U.S. defense tech companies and some federal agencies to direct employees to stop using Anthropic’s Claude model for defense work. Venture and defense firms with DoD contracts are actively moving workloads off Claude and seeking replacements; Palantir and other contractors are expected to remove Anthropic technology from sensitive supply chains. Anthropic previously entered DoD networks via a Palantir partnership and secured a $200 million contract to deploy Claude in classified environments. OpenAI said it reached terms with the DoD and clarified restrictions on domestic-surveillance uses. Anthropic disputes the authority of Defense Secretary Pete Hegseth’s directive and may pursue legal options if the designation becomes formal.
A U.S. government supply-chain designation affecting a major LLM vendor can force DoD contractors and federal agencies to remove or replace Anthropic’s technology, causing operational disruption across defense AI deployments and prompting vendor migrations.
Track Anthropic Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- The Trump administration announced a supply-chain risk designation for Anthropic, publicly stated by Defense Secretary Pete Hegseth.
- Multiple defense tech companies have instructed employees to stop using Anthropic's Claude for defense contracts and are replacing it with other models.
- Anthropic entered the Department of Defense ecosystem in late 2024 via a partnership with Palantir and won a $200 million contract to deploy Claude in classified networks.
- OpenAI announced it agreed terms with the DoD and amended contract language to prohibit intentional use for domestic surveillance of U.S. persons.
- Federal agencies including the Defense, Treasury, State, and Health and Human Services Departments have directed employees to move off Claude; President Trump said agencies have six months to phase out the technology.
Connected Companies & Entities
6 Entities mappedOntology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Palantir Sticks with Anthropic's Claude Despite Pentagon Concerns
Palantir CEO Alex Karp said Palantir continues to use Anthropic’s Claude models in its products even after the U.S. Department of Defense designated Anthropic a supply‑chain risk. Speaking at Palantir’s AIPcon 9 in Maryland during an interview with CNBC’s Seema Mody, Karp said the Pentagon’s designation has not yet resulted in a full phase‑out of Anthropic from defense workflows and noted the Department of Defense is still using Claude to support operations in Iran. Karp added that Palantir’s products are integrated with Anthropic today and that the company plans to add other large language models in the future as the situation evolves. The comments clarify Palantir’s current vendor integrations amid the DoD’s recent policy action.
Pentagon Flags Anthropic's Claude as Supply Chain Risk
U.S. Defense Department CTO Emil Michael told CNBC that Anthropic’s Claude AI models were designated a supply chain risk because their policy preferences are baked into the models and could "pollute" the defense supply chain, potentially delivering ineffective equipment to warfighters. The designation, the first publicly applied to an American AI company, requires defense contractors and vendors to certify they do not use Claude in Pentagon work. Anthropic has sued the Trump administration, calling the action "unprecedented and unlawful," and said the designation jeopardizes hundreds of millions of dollars in contracts. Michael said the move is not intended to be punitive and that only a small fraction of Anthropic’s business comes from the U.S. government. Anthropic, founded in 2021 by former OpenAI researchers and executives, is known for its Claude family of models and enterprise sales, including previous Department of Defense engagements.
US Military Sticks with Claude Amid Contractor Exodus
Anthropic's AI models (branded as Claude) remain in active use by the U.S. military for targeting decisions during the U.S.-Israel strikes on Tehran, even as many defense contractors and subcontractors move away from the company. President Trump directed civilian agencies to stop using Anthropic products and gave the company six months to wind down Department of Defense (DoD) engagements, but the directive has not fully curtailed DoD use. Reporting in The Washington Post says Anthropic systems were used alongside Palantir’s Maven to suggest and prioritize hundreds of targets with precise coordinates. Defense contractors including Lockheed Martin and multiple portfolio companies associated with J2 Ventures are replacing Anthropic models, citing defense-use concerns, while Secretary of Defense Pete Hegseth has threatened a supply-chain risk designation but no formal action has yet been taken.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.