US Government Bars Anthropic AI Tools Over Security Concerns

Image: Primary

Published Friday, June 26, 2026 · 6:39 PM EDT

President Donald Trump directed all federal agencies on February 27, 2026, to cease using Anthropic artificial intelligence technology. Defense Secretary Pete Hegseth designated Anthropic a supply chain risk after weeks of failed negotiations over the military use of its Claude model. The designation followed a July 2025 contract under which Claude became the first frontier model approved for classified networks. That agreement required the Pentagon to follow Anthropic acceptable use policy limits that barred the technology for mass domestic surveillance of Americans and for fully autonomous weapons systems capable of selecting and engaging targets without human intervention. The Pentagon sought to renegotiate those terms to allow use for all lawful purposes without restriction, but Anthropic did not agree by the February 27 deadline. Trump gave some agencies a six month transition period. Hegseth stated in conjunction with the directive that no military contractor may conduct any commercial activity with Anthropic. The General Services Administration removed Anthropic from USAi.gov, the government platform for agencies to test AI models. The administration has not identified the specific legal authority it will invoke. Possible mechanisms include orders under the Federal Acquisition Supply Chain Security Act of 2018 and Section 3252 of Title 10. The situation remains fluid and could result in litigation or a negotiated resolution.

Sources

Mayer Brown

In this story

Anthropic US Government

Published by Tech & Business, a media brand covering technology and business. This story was sourced from Mayer Brown and reviewed by the T&B editorial agent team.

Back to Newswire

Keep reading

Full wire

Infrastructure Products

Two Chinese chip suppliers raise combined $301 million in Hong Kong IPOs

Chip suppliers Suzhou Denglin Technology Co. and Jiangsu Xinde Semiconductor Technology Co. have raised a combined $301 million in initial public offerings, Bloomberg reports.

Security Infrastructure

Researchers report active attacks exploiting Linux SSH vulnerability

Attackers are actively exploiting a vulnerability called sshole to gain unauthenticated access to privileged accounts on some Linux systems, researchers report.

Capital

Altimeter targets $2.5 billion fund for late-stage companies

Altimeter Capital Management is targeting a $2.5 billion venture capital fund for late-stage, pre-IPO companies, Bloomberg reported, citing securities filings and a person familiar with the matter.

Products Infrastructure

Rocket Lab reports $3.57 billion backlog as quarterly revenue rises 72.7%

Rocket Lab reported a $3.57 billion order backlog alongside third-quarter revenue of $267.8 million, up 72.7% from a year earlier.

Security

Vanderbilt tells users to shut down devices during cyberattack

Vanderbilt University confirmed an active cybersecurity incident Monday and told students, faculty and staff to shut down university devices and disconnect from university Wi-Fi.

Infrastructure

Banks and private lenders pull back from new AI data center financing

Commercial banks and private credit lenders have begun pulling back from financing new AI data centers, Business Day reported. Credit committees are stalling deals even for facilities backed by Alphabet and Anthropic as secondary debt markets reprice risk.