Breaking

Lutnick and Pentagon CTO Reportedly Met Anthropic Executive Tom Brown as AI Safety Fight Intensifies

September 15, 2026 at 18:38 EDT

Foundation Models
AI Agents
Policy & Regulation

Commerce Secretary Howard Lutnick and Pentagon chief technology officer Emil Michael reportedly met virtually with Anthropic executive Tom Brown to discuss AI safety risks, bringing a public policy fight over frontier AI directly into talks between the company and senior U.S. officials. The reported meeting comes as Anthropic, the maker of Claude, is pressing for stronger safeguards while parts of the Trump administration remain divided over whether safety rules would reduce risk or slow U.S. competition with China.

Sep. 2026 · AI safety and U.S. policy

Anthropic's safety fight reaches Washington

Senior U.S. officials reportedly discussed AI safety risks with Anthropic as the company pushes for stronger safeguards and the Pentagon seeks broader access for defense missions.

6-12

months in Amodei's risk window

months covered by Anthropic's report

harm areas documented

months for Pentagon Claude phaseout

Threat report scope, drawn to scale

8 monthsactivity reviewed

7 areasharm categories

The documented risks span cyber operations, surveillance, influence operations, scams, biological misuse, conventional weapons and illicit distillation.

Safety case

Supporters see independent evaluation and limits on fully autonomous weapons as necessary checks on increasingly agentic systems.

Policy concern

Critics warn that regulation could entrench dominant AI labs or slow U.S. competition in defense and geopolitics.

The policy path

Frontier models

More capable agents

→

Misuse evidence

Cyber, weapons, surveillance

→

Government choice

Safety limits or broad access

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.