Breaking

Bernie Sanders Says AI Risk May Exceed Nuclear Weapons as He Pushes Superintelligence Ban

September 17, 2026 at 00:15 EDT

Policy & Regulation
Foundation Models
AI Agents

Sen. Bernie Sanders escalated Washington's AI safety debate by saying the potential danger from artificial intelligence is "probably greater" than nuclear weapons, while preparing legislation to ban artificial superintelligence and pause advanced AI development. The Vermont independent is proposing penalties of up to 20 years in prison for violations, putting his plan among the most aggressive AI restrictions now being discussed in Congress.

Sept. 15, 2026 · U.S. Congress

AI Risk Enters Nuclear-Tier Politics

Bernie Sanders says AI's potential danger may exceed nuclear weapons as he backs a superintelligence ban, a development pause, and criminal penalties for violations.

20 years

Maximum proposed prison term

1,000+

AI agents cited in the July episode

$32B

Prior Senate AI funding recommendation

Penalty Comparison

20 years

Banned superintelligence violations

20 years

Illegal nuclear-weapons development comparison

Each block represents two years; the proposed AI penalty is framed at the same length as the nuclear-weapons comparison.

What the Bill Would Do

Ban superintelligence

→

Pause advanced AI

→

Create federal review

Supporters' Case

Frontier systems are becoming more agentic, and voluntary company safeguards may not be enough if control failures grow more severe.

Critics' Concern

A broad pause or ban could be hard to define, difficult to enforce, and potentially costly for smaller competitors.

Continue reading

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