Reporting archive
Evidence on AI and work
-
AFL-CIO makes AI the 2028 campaign litmus test
Organized labor just turned AI policy into a campaign gatekeeper—and 2028 hopefuls will either meet worker standards or fail the test.
-
7,000 reassigned, spans shrink, Meta rebuilds its management graph
Zuckerberg admits Meta’s AI pivot overshot, pauses broad layoffs, and pivots from buying clusters to rebuilding the org that ships them.
-
Walmart’s 2.1 million associates get an AI literacy baseline
Walmart is wiring AI into design, logistics, and store ops with a bold promise—train everyone, don’t trim headcount—and a nationwide OpenAI-backed certification to put blue vests in charge of the outputs.
-
Sanders seeks 50% public stake to veto mass automation
Sanders wants the off switch on the cap table: a public 50% stake with veto over job‑erasing AI deployments, plus a pause on new data centers until their footprint is counted.
-
Anthropic puts $200 million behind unemployment-triggered job protections
Anthropic just put $200M behind measuring AI’s job shocks—and paired it with a trigger-based plan that pays when the market won’t.
-
JPMorgan lets autonomous agents run for an hour or two
JPMorgan is graduating from chatbots to long‑running, permissioned agents—complete with audit trails—redesigning white‑collar work before it redesigns headcount.
-
UK debuts experimental Work Hub AI job coach
Britain didn’t just talk AI—it shipped a 24/7 public job coach, and the next three months will reveal whether it turns layoff into interview time, not rhetoric.
-
Data center payrolls grow; org charts refuse to autocomplete
Billions are flowing into AI, but the only obvious hiring is pouring concrete—because until costs fall and governance settles, “adoption” without real integration just adds steps, not pink slips.
-
UK launches £20m alliance to design paid AI beginner roles
The UK just turned “train and pray” into paid guarantees, rebuilding the missing first rung of the AI economy with contracts, not certificates.
-
Walmart’s 4.95% vote keeps AI’s labor uncounted
A 4.95% vote at the nation’s largest employer signals investors prefer AI speed over standardized worker-impact metrics—and gives boards cover to keep the numbers in the dark.