2026-09-01 — A second, much larger Kaiser union takes AI-staffing language to the picket line

• California Nurses Association/National Nurses United — the 25,000-nurse general-RN union at Kaiser, distinct from the ~2,400-clinician NUHW mental-health bargaining unit this thread has tracked since March — held informational pickets at 22 Kaiser hospitals across California on 2026-09-01, demanding contract language requiring new technologies be “tested, regulated, and subject to nurse input” before deployment, alongside safe-staffing demands. Hospitals included Oakland, San Francisco, Santa Clara, San Jose, Walnut Creek, Fremont, Redwood City, Sacramento, Fresno, Los Angeles, and eleven more. CNA/NNU is in early-stage bargaining, separate from NUHW’s; a Kaiser spokesperson said its technology is meant to “support nurses rather than replace their clinical judgment” and that these principles are “already written into” its CNA contract — a claim not yet tested by anything in this round. This widens the “whether other health systems/unions face the same” watch line to a second bargaining unit inside Kaiser itself, an order of magnitude larger than NUHW’s, applying pressure “early in the bargaining process, before staffing and AI protections are settled” per NNU’s own framing. (National Nurses United, KRON4)

• San Francisco’s Board of Supervisors unanimously adopted a resolution on 2026-09-01 calling on Kaiser Permanente to withdraw its “Terrible Three” contract demands and negotiate in good faith, escalating past July’s procedural “heard and filed” hearing into a formal, on-the-record city position. Supervisor Chyanne Chen, who introduced it, said “how we’re taking care of our patients and workers, especially with emerging technology, is more important now than ever”; NUHW President Emeritus Sal Rosselli said “Kaiser officials should be eager to work with their clinicians to improve services, but instead they want to sideline them.” Kaiser told KALW it “does not use AI to make clinical decisions or replace licensed clinicians,” while the same article reports therapists already seeing AI in triage practice. No mediation dates have been scheduled since four days of talks in August ended without progress. ⚠️ Out of window — the vote happened 09-01, before this run’s W1 (2026-09-03 15:30 ET) begins — but it directly resolves an item the 09-03 digest explicitly flagged as “seen in the buffer, not resolved to a citable source this run — to verify next run”: this is that verification, now primary-sourced. (NUHW, KALW)

2026-08-30 — AB 2575’s second attempt needs a second reading first, and that hasn’t happened yet either

• Checked directly against the Legislature’s own bill-history record: because AB 2575 was amended when reconsideration was granted on 08-28, it went back to second reading rather than straight to a repeat third-reading vote — a procedural step the 08-29 entry didn’t have. As of this check (Sunday 08-30, ~10am ET) it sits on the Senate’s 08-30 second-reading file, “in floor process,” with no second floor-passage vote yet recorded. One day remains before the 08-31 adjournment for that second reading and the repeat vote both to happen. (AB 2575 bill history)

2026-08-29 — AB 2575 nearly fails, and a second health system’s clinicians turn up with the same complaint

• AB 2575 — the bill Kaiser and the California Hospital Association are fighting — failed its Senate floor vote 08-28 (18-10) and was immediately granted reconsideration (40-0), leaving it alive for one more attempt before the legislature adjourns 08-31. Verified against the Legislature’s own bill-history record. Full detail cross-logged on The Reckoning. (California Legislature — AB 2575 bill history)
• Kaiser’s own spokesperson tied the AI-triage dispute directly to active NUHW contract bargaining for the first time on the record, telling The American Prospect its bargaining proposals are “focused on ensuring our members have timely access to high-quality mental health care, supporting our clinicians in delivering excellent care” — while separately maintaining the February 2025 DMHC settlement “did not involve AI.” Confirms this thread’s framing that the fight is running through labor bargaining, from Kaiser’s own mouth rather than inference. (The American Prospect)
• A second health system surfaced with the same pattern: Rogers Behavioral Health (Wisconsin) partnered with AI vendor Limbic in late 2024 for intake/triage, and two named Rogers clinicians describe the same inappropriate-placement and reduced-risk-screening complaints as Kaiser’s. Rogers’ West Allis workers voted 54-4 to join NUHW in April 2026, and Rogers has not yet agreed to bargain — the same union, organizing a second employer, over the same AI-triage complaint. This is the “whether other health systems face the same” watch item resolving from hypothetical to named. (The American Prospect)

2026-08-27 — Nine triage clinicians to three, as algorithmic routing takes over

• An American Prospect investigation reports Kaiser Permanente’s Walnut Creek clinic cut mental-health triage staffing from nine clinicians to three over three years while leaning on automated and algorithmic triage, with clinicians describing patients going uncontacted or routed to a lower level of care than assessed need warranted. NUHW filed a second California DMHC complaint on 2026-07-20 over Kaiser’s web-based “e-visit” triage tool. The substitution is the claim that matters, not the automation. A triage tool arriving is unremarkable; a triage tool arriving as six clinician posts leave is a staffing decision a regulator can act on, and it is the specific shape this thread has been waiting to see documented. (The American Prospect, SF Standard, the July complaint)

2026-08-21 (late catch, added 2026-08-23) — AB 2575 confirmed real, and it moved on the Senate floor

• California AB 2575 exists, the bill number reported alongside the Kaiser investigation is correct, and it was read a second time and amended on 2026-08-21. Verified directly against the California Legislature’s own bill record, not the coverage: author Assemblymember Liz Ortega, subject “Health care services: artificial intelligence,” amending the Business and Professions, Civil, Health and Safety, and Labor Codes. Status is in the Senate floor process, ordered returned to second reading. The 08-22 digest specifically held this pending a check against the legislature’s own record; that check is now done and the number is confirmed. (California Legislature — bill record, primary)
• What the bill would do: protect health-care workers from retaliation for overriding an AI system’s clinical recommendation, require transparency about AI use and risks in care, and shift liability for AI-caused patient harm onto AI developers and facilities. Kaiser and the California Hospital Association oppose it. This is the legislative form of exactly the conflict this thread tracks — a clinician’s right to override the model, written as statute. (Capital & Main)

2026-08-18 (late catch, added 2026-08-23) — The staffing number behind the algorithm: nine triage clinicians to three

• Capital & Main published an investigation finding that a Kaiser Permanente Northern California facility’s mental-health triage team shrank from nine clinicians to three as work shifted to an automated e-visit screening tool in place since 2023. Based on interviews with more than a dozen therapists, clinicians, academics and advocates in California and Wisconsin, it reports weeks-long care delays, missed self-harm and suicide-risk flags, and patients rerouted to apps and call centres rather than clinicians. This is the first hard staffing ratio this thread has had — the union complaint it builds on alleged the practice; this puts a number on it. (Capital & Main, primary)
• It extends, rather than resets, the NUHW regulatory complaint already on this thread — filed roughly three weeks earlier (late July, first reported by CalMatters) alleging the same e-visit tool triages without clinician review in violation of state law.

2026-08-14 — Mediation window’s scheduled last day closes with no outcome posted either side

• The mediation window Kaiser itself described as “August 11 to 14” reaches its scheduled end today, and a direct check of both parties' own channels found nothing new: Kaiser’s labor-relations page still shows its August 11 “next steps” message as the newest post, and NUHW’s news page hasn’t moved past July 27. That August 11 message restates Kaiser’s offer in full — 17.5% total wage increase over four years — and confirms Kaiser’s proposed joint AI committee for NUHW is still on the table, explicitly modeled on the existing AI framework agreement Kaiser already has with the Alliance of Health Care Unions (recognized by MIT Sloan as “the first comprehensive effort by a major healthcare employer and its union partners to integrate worker voice into all phases of AI decision-making”). Echoes the same silence-not-signal pattern that produced a passed-silent flip on the mediation’s opening day; unlike that entry, no upcoming.yaml expectation currently covers this specific close date. (Kaiser Permanente — NUHW bargaining, NUHW)

2026-08-11 — Mediation began today and closed the digest-day confirmed passed-silent

• Mediation between Kaiser and NUHW began today, and by the time the digest-day closed, a direct check of both parties’ own channels — run three separate times across the day — found no readout at all. NUHW’s news page never moved past its July 27 post; Kaiser’s own mediation page (about.kaiserpermanente.org) never added anything past its original June 11 announcement that the parties agreed to mediate. The upcoming.yaml ledger entry (kaiser-nuhw-mediation-0811) is now formally flipped to passed-silent — the loud outcome this ledger is built to flag rather than an assumed gap: a scheduled mediation session came and went with neither side saying anything publicly, and this same dispute had a mediation round go silent before (2025 precedent). Watch next for whether either side breaks silence in the days after, and whether SB 903’s Assembly path (appropriations hearing 08-13) becomes the venue where the AI-replacement language actually gets decided instead. (NUHW; Kaiser labor relations)

2026-08-07 — NUHW’s fight now named alongside SB 903 in national press

• An AP wire story on SB 903 (“As AI ’therapists’ dish out advice, California lawmakers try to set some limits”), syndicated nationally, names NUHW as a backer of the bill — the first time this thread’s labor fight and Chatbot Bans’ legislative track have been tied together in one general-audience story rather than covered separately. No new fact about the Kaiser dispute itself; logged here because it’s the first public linkage of the two threads’ stories. Full detail on the wire story lives on Chatbot Bans' own timeline (2026-08-07 entry). Published evening of 2026-08-06 ET, ahead of the 2026-08-07 digest window. (News From The States, AP wire)

2026-08-05 — The mediation date the thread had been missing: August 11

• NUHW confirms mediation with Kaiser starts August 11 — resolving the “no firm date yet” gap this timeline has carried since 07-23 (Kaiser’s own labor-relations page had only said NUHW wasn’t available until August, with no date attached). NUHW frames the deadlock around three demands: AI-replacement language (“Kaiser wants to be able to replace therapists with A.I.”), layoff/outsourcing provisions, and a “sole discretion” clause over BH clinical operating models. Kaiser’s own account of its AI proposal, by contrast, describes only a joint clinician/union/management advisory committee on AI use — the framing gap between the two sides' descriptions of the same proposal is itself part of the dispute. (NUHW, 07-27; Kaiser labor relations)
• This is a gap catch, not a new event — the NUHW source is dated 07-27 and slipped past prior passes (last_seen moved to 08-04 without an entry capturing it); logged now because it directly resolves the thread’s open “no firm date” watch item.

2026-07-29 — DMHC confirms it’s investigating the algorithm-triage complaint

• The National Union of Healthcare Workers filed a complaint with California’s Department of Managed Health Care alleging Kaiser used an algorithm, not clinicians, to triage mental-health e-visit patients, and the department confirmed it is investigating. (Almanac News / CalMatters)
• This is the first formal regulatory complaint on this specific practice, and it moves the question from what a chatbot tells a consumer to what an algorithm does inside a licensed care system.

2026-07-23 — “Unacceptable”: the hearing gets a second wave

• Coverage of Tuesday’s SF Board of Supervisors hearing crested a day late — Kaiser no-showed; Supervisor Walton called it “extremely disrespectful,” Supervisor Chen called the testimony “infuriating, concerning and unacceptable.” NUHW’s “Terrible Three” framing crystallized: AI expansion without a no-replacement guarantee, layoffs paired with outside-contractor referrals, and Kaiser taking sole control of BH clinical policy. (KQED)
• No firm mediation date yet — per Kaiser’s own labor-relations page, NUHW indicated it’s “not available until August,” so Tuesday’s hearing pressure hasn’t produced a locked session date.

2026-07-22 — Complaint public, mediation agreed

• NUHW’s formal DMHC/DOL complaint went public — 29 pages alleging the E-Visit tool generates care recommendations in ~2 seconds despite a claimed 4-hour clinician-review window, and that only high-school-diploma clerical staff conduct the screenings. (SF Standard)
• The SF Board of Supervisors hearing happened — Kaiser skipped it — the board unanimously passed a resolution demanding “full parity for mental health patients”; both sides have now agreed to pursue mediation, with Kaiser denying the AI/clinical-decision allegations. (KALW)

2026-07-21 — Surfaced (critic catch)

• 2,400 Kaiser BH clinicians kept up the pushback on AI in care decisions — with San Francisco’s Board of Supervisors holding a hearing; BHB’s lead that day, caught by the coverage critic (the sweep missed it — Kaiser wasn’t on the watchlist; critic-add applied same day). (Behavioral Health Business)

← Backstory (crawl 2026-07-22 → artifacts/findings/kaiser-ai-clinician-backlash-2026-07-22.md)

2026-07-20/21 — Complaint + hearing (the week the digest caught)

• NUHW filed a 29-page formal complaint to DMHC + US DOL — the “E-Visit” auto-triage tool and unlicensed TSR phone triage as illegal unlicensed triage; instant unreviewed AI referrals documented with dated screenshots. (complaint PDF)
• The SF hearing was a formal Committee of the Whole (File 260747, Sup. Chen; clinician + expert panels; Kaiser declined to present; procedural “heard and filed”). (SF Board agenda)

2026-03-18 — Behavioral health’s first union strike over AI

• ~2,400 clinicians struck for one day (92% authorization; ~23k CNA nurses in solidarity) on the ULP charge over Kaiser’s unilateral triage overhaul — days after Kaiser’s $31M DOL parity settlement (02-10). (NUHW)
• Bargaining then stalled on the “Terrible Three” — AI replacement at will, outsourcing, “sole discretion” over clinical decisions; Kaiser’s 06-04 counter was a non-binding quarterly AI committee (“toothless” — NUHW). Mediation set for ~August.

2023→2025 — The systems and the rap sheet

• The AI at issue is specific: late-2023 “E-Visit” questionnaire (partial PHQ-9/GAD-7) auto-generating referrals + Jan-2024 unlicensed TSR phone triage — replacing licensed triage clinicians. (KQED)
• Regulators were already on the field: $200M DMHC settlement (2023); DMHC’s 02-2025 survey citing failed suicide-risk screening in triage; Kaiser signed “AI will assist, not replace” with its SoCal union then deleted it from the NorCal proposal. (DMHC survey)
• The legislative echo: NUHW-co-sponsored SB 903 (bar AI-only screening by health plans) passed the CA Senate — cross-ref Chatbot Bans.