Marketplace All-in-One What's behind the Anthropic-Pentagon feud?
Feb 26, 2026
Kion Vestensen, a senior researcher at Freedom House focused on digital rights and democracy, discusses Anthropic's shift away from strict AI safeguards and pressure from the Pentagon over model use. He also explores age-verification rules, risks from ID and facial-scan systems, and privacy-preserving alternatives. Short, clear conversation on safety, surveillance, and possible fixes.
AI Snips
Chapters
Transcript
Episode notes
Anthropic Moves From Guardrails To Goals
- Anthropic shifted from binding guardrails to non-binding safety goals after other companies and governments failed to coordinate.
- The company had promised stricter safeguards if Claude could help build weapons but rescinded hard rules when peers didn't follow and regulators didn't align.
Pentagon Pressures Anthropic Over Military Use
- The Pentagon is pressing Anthropic to remove constraints so its AI can be used rapidly in military scenarios like time-sensitive weapons firing.
- Defense Secretary Pete Hegseth reportedly gave an ultimatum linking a $200 million contract and may use supply chain designations or the Defense Production Act as leverage.
Avoid Bulk ID Uploads For Age Verification
- Avoid blunt, large-scale ID uploads for social media age checks because they centralize sensitive data and increase breach risk.
- Prefer developing third-party privacy-preserving tokens that verify age then delete IDs, though these solutions need time and funding.