House Speaker Mike Johnson (R-La.) said Sunday on CNN's "State of the Union" that Congress should not take the lead in regulating artificial intelligence (AI) safety and that an immediate congressional-imposed moratorium would be the wrong approach.
Why it matters
Calls intensified this week for AI companies to adopt safety measures and slow their development. The four largest AI labs signaled support for slowing model development, and some lawmakers urged canceling the House recess so representatives could return to Washington and pass legislation on AI safeguards.
Johnson's position
Johnson rejected the idea of congressional intervention as the primary solution, saying companies could pause their work if needed and warning against an emergency moratorium. He said, "We have to resist Congress jumping in and imposing some sort of emergency moratorium. It's got to be done right. Safely. Wisely."
He argued that a complete halt to AI development would pose national security risks because "China will overlap us," and proposed a meeting including President Trump, members of Congress, and top AI executives to discuss safeguards.
Johnson said he would prefer that AI company leaders assume leadership and responsibility for the safety of their products rather than pushing Congress to impose rules. He told company executives they must "do your part" and acknowledged their corporate responsibility for product safety, noting these firms are pushing the envelope and must be accountable.
Related views and context
Johnson's remarks echoed those of former White House AI czar David Sacks, who late Saturday said major labs should voluntarily pace development without government intervention.
CNN host Jake Tapper noted that several AI leaders agreed with a 3,800-word essay published Saturday by Anthropic's Dario Amodei, which proposed slowing AI development and adopting more safety measures.
Johnson expressed skepticism that companies share a consensus on appropriate guardrails: he said they "all have very different ideas on what the guardrails should be. There's no consensus among them." He also suggested that Congress is less qualified than frontier developers to understand every technical detail.
Further considerations
Johnson urged Americans not to panic over doomsday scenarios tied to the technology, while acknowledging that rank-and-file members of Congress have been urging action, as reported last week by Axios' Andrew Solender.
He argued the new technology should be handled as past innovations were: responsibly, without smothering American innovation. "We have to do both things simultaneously," he said.
Summary
The debate centers on whether AI safeguards should come from congressional action—potentially through rapid emergency rules—or from industry-led pacing and self-regulation. Recent developments, including major labs endorsing slower development and Dario Amodei's lengthy essay, have increased pressure for action, while Speaker Johnson cautioned against hasty government intervention and emphasized the need for corporate responsibility and coordinated discussions among political leaders and tech executives.
