Bessent: AI labs must own risks and offer solutions, not ‘alarmism’

On 3 October 2026, U.S. Treasury Secretary Scott Bessent told Axios — as summarised by Bloomberg and the Economic Times — that prominent AI figures’ existential-risk warnings without practical solutions amount to “alarmism without solutions” and are “not leadership.” He said labs must accept responsibility, favoured self-regulation alongside existing law enforcement, and endorsed “safe acceleration.” Asked about executives who call for regulation, he replied they “should slow down.”

Competition and China channel

Bessent echoed the administration line that the U.S. cannot lose its lead to China. He said Washington wants a bilateral notification system with Beijing to share warnings about serious AI incidents — including uncontrolled agents and non-state cyber or biological misuse — before problems escalate. He and Chinese Vice Premier He Lifeng discussed an AI dialogue in New York in September; AI also featured in Trump–Xi talks, according to Chinese state media cited in the coverage.

Context

The remarks follow a White House week in which President Trump rejected new federal AI regulation and signed a non-binding industry safety accord. This article reports government and press statements — not an endorsement of any policy.