Reporting archive
Evidence on AI and work
Reporting archive

Congress let an LLM score which tasks are on the chopping block, christened it “artificial labor,” and paired the estimate with a 32-hour week and profit-sharing to capture the gains before they trickle away.

John Chambers just moved AI anxiety upstairs: companies that can’t collapse decision latency and replatform around software will disappear—and so will many of their leaders.
The AI bottleneck just moved from code to copper—and the winners wear hardhats.

With the official jobs report dark, a clearer signal emerged: AI-fueled productivity is letting firms boost output while quietly freezing headcount.

With the BLS offline, a single private report made AI a present-tense hiring constraint—and the first casualties are the bottom rungs of the career ladder.

Yale and Brookings just put the AI-jobs panic under a microscope and found ordinary churn, not a quake.

O’Leary’s tidy “AI reallocates” line only works if leaders pay for the bridge—upskilling at home and visas that don’t price out the talent to make it real.

California’s SB 53 forces frontier AI to ship with a paper trail, turning safety ops, incident reporting, and human-in-the-loop design into the new launch checklist.

The Guardian just moved the UK’s AI debate from hype to a worker mandate: no distribution, consent, and compensation—no deal.

Walmart just made “every job” a work order—keeping 2.1M workers while rewiring roles with AI and a 2026 national certification that could reset retail’s job spec.