Reporting archive
Evidence on AI and work
Reporting archive

In May, U.S. employers didn’t just talk about AI—they checked it on the layoff form, with 40% of cuts carrying the label.

BCG’s latest study shows AI has turned many roles from doing to directing—creating newfound hours that mostly leak away without a bold workflow rewrite.

When a $2.2T bank lets AI write first and makes humans prove what survives, the winners are editors, not typists.

While peers freeze hiring, Cognizant is onboarding 20,000 grads, formalizing AI-operator roles, and replacing token tallies with outcome guarantees.

A single payroll series just hijacked the AI layoff narrative, pointing to an implementation economy that’s hiring faster than automation can subtract.

Connecticut just turned layoff spin into evidence by forcing employers to declare—on the record—whether AI helped cut the jobs.

AI’s biggest hiring wave is under hard hats—and it vanishes the moment the ribbon is cut.

JD just turned ‘AI did it’ into a non-reason for layoffs, betting it can redeploy humans as fast as it automates.

A Fed governor just put a timeline on AI—job losses before gains—turning the transition into a sequencing test for labor, inflation, and boardrooms.

The OECD just turned AI “exposure” from vibes into a living map, showing clerical work nearly on the line and embodied, high-accountability roles still far away.