Breaking

Trump Advisers Met Anthropic Executive Tom Brown as AI Safety Debate Intensifies

September 16, 2026 at 01:16 EDT

Policy & Regulation
Foundation Models
Security

Senior Trump administration officials held a virtual meeting with Anthropic's chief compute officer Tom Brown on Sept. 15 to discuss AI safety risks, according to Bloomberg reporting republished by NDTV Profit. The discussion brought together Brown with Pentagon Chief Technology Officer Emil Michael and Commerce Secretary Howard Lutnick at a moment when frontier AI oversight has become a live political fight.

Sept. 15, 2026 · Anthropic

AI safety reaches the cabinet table

Senior Trump officials met Anthropic executive Tom Brown as the company pushed a plan to pace frontier AI development while Washington remained divided over regulation.

Sept. 15

meeting date

pacing steps

6-12

months cited

harm areas

Risk window visualized

6 monthslower end

12 monthsupper end

Amodei's warning centered on a 6-12 month window in which more capable misaligned agent swarms could pose far larger cyber risks.

Pacing framework

1. Embedded evaluators

Outside reviewers get employee-like access.

→

2. Democratic coordination

Frontier labs align on standards and limits.

→

3. Global coordination

Governments try broader agreements.

Supporters

Embedded evaluators and common standards could make safety claims easier to verify before stronger models are deployed.

Critics

Voluntary coordination could favor incumbent labs and raise concerns about regulatory capture.

Continue reading

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