Governance & Safety

The left is splitting over which AI risks count

The fight is not simply between AI supporters and critics. WIRED reports that critics disagree about which risks deserve attention. Treating either side as fake leaves a gap: current harms need concrete rules, while frontier risks need outside tests, incident reports, audits, and limits on dangerous capabilities.

AI Agency
Read original source

What the source reports

Caroline Haskins reports on a split within the political left over AI risk. One camp says talk of human extinction hides harms that people face now. Those harms include threats to workers and privacy. They also include surveillance and the effects of large data centers. This camp worries that leading AI firms gain from the focus on doom. Another camp says both sets of risks need rules. The source calls for worker, privacy, buying, and surveillance rules now. It also names outside tests, incident reports, audits, and legal duty for leading systems. This is political reporting. It does not show that either plan has been tried or passed.

Original source

Title
The Leftist Split Over AI Doom
Author
Caroline Haskins
Publication
WIRED
Date
Friday, September 18, 2026