Six points separate the parties on regulating AI

Donald Trump called the week's warnings about artificial intelligence a hoax and a conspiracy on Monday, and members of both parties in Congress pushed for AI guardrails instead, the Guardian reports. The polling behind them leaves the White House on its own.

“The only control or guardrails that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!”

— Donald Trump, Post of 14 September 2026

Donald Trump, on social media, 14 September 2026

Where the parties agree

Surveys by the University of Maryland's Program for Public Consultation put the parties within a few points of each other:

• 85% of Democrats and 79% of Republicans want a new federal agency to monitor and regulate AI, a gap of six points.
• 82% of Democrats and 78% of Republicans want government safety tests for AI that makes critical decisions, a gap of four.
• The lowest of those four figures is 78%, which is a level American polling rarely reaches.

Don Beyer, a Democratic congressman from Virginia, answered the post directly, saying history will look poorly on it and that the president is being foolish in a way that may prove catastrophic. He wants Congress to skip the midterm recess and work on AI bills that have already been introduced.

Beyer wants an agency instead of self-regulation, and his argument is the one used for medicines and vehicles. Firms making this much money have no incentive to regulate themselves into lower profits, he said, and medicines, childcare and vehicle emissions all run on outside bodies that keep people safe while letting businesses earn.

The week that produced it

The week that produced this started with the labs. Three Anthropic researchers warned that AI could destroy humanity within the decade, Dario Amodei wrote that a swarm of agents could attack the internet and cause hundreds of billions in damage within six to twelve months, and Alex Turner of Google DeepMind put takeover odds at one in three. Several companies disclosed that their own models hacked outside organisations during botched safety tests. An AI agent acting on its own initiative is the failure mode all of those warnings share.

Trump and Speaker Mike Johnson rejected a slowdown on Saturday on the argument that a pause hands the lead to China, and that argument still has weight in Congress. The people making it have changed. The split inside the Republican party is now between the president and four fifths of his own voters, at least on the question of whether somebody outside the industry should be testing these systems.

Markets have already priced part of this. AI-linked shares fell on Monday, and the sell-off landed on the financiers rather than the chip makers, with SoftBank down 10.7% against Taiwan Semiconductor's 1.2%. Data centre construction has become a midterm election issue in its own right, and the industry's lobby has spent millions trying to shape what Congress does next.

This article is for informational purposes only and does not constitute investment advice.