Cycle on Oct 19, 2026

I'll start by reviewing the topic's open threads and existing notes, then pull market data and fresh evidence.

I have the full picture: 10 threads including two watches that may have fired (Tesla's Washington FSD insurance program, fire date Sept 15; and the Q3 insurer-earnings watch, fire date Oct 15). Today is Oct 19 — Progressive's September results should be out, Tesla reports Q3 on Oct 21, and the NHTSA Cybercab deadline is Oct 30. First, market data on the named insurers and AV operators.

Insurer data pulled. Now the remaining named players: Root, Kemper, Aurora, Berkshire (GEICO).

BRK.B ticker failed (will try hyphenated form). Now web evidence — Progressive's September results, the Tesla Washington insurance watch, and Waymo/Tesla scaling news.

Progressive's September/Q3 results are out (Oct 14) — NPW growth slowing to +2% monthly, stock fell 4.2% on the print. Let me get the full release detail and search the AV-operator side.

Strong material: Progressive's first sequential personal-auto PIF decline since Nov 2023, Waymo opening three cities + Toronto, Tesla's Cybercab fleet 8x in 30 days. Now the Waymo blog primary source, the pending Washington Tesla-insurance watch, and the new NHTSA probes.

Waymo's blog gives me primary-source safety data (270M miles, 82% fewer injury crashes) and the Toronto/three-city expansion. Now let me verify the Tesla Washington insurance watch (fire date was Sept 15) and check the Lemonade and AI-agent tripwires.

The Washington program was still pending approval as of that article (filed, tentative Sept 1 launch). Lemonade's Oct 8 HW3/FSD-v14-Lite expansion fires its ongoing watch. Let me verify the Lemonade primary item, the Travelers agentic-AI commentary, and make one final check on the Washington launch.

I have all the evidence gathered and verified — now persisting it. Writing the four finding updates (Progressive PIF erosion signal, Waymo scale/safety, Tesla robotaxi/insurance, Lemonade FSD pricing).

First note saved. The second call was malformed (missing parameters) — rewriting the remaining three notes properly now.

All four findings persisted. Now recording the two watch events that fired this cycle (Progressive's Q3 read; Lemonade's rollout) and registering three new forward-looking watches.

Cycle complete. Summary of what this cycle found and persisted: **Findings (4 note updates):** 1. **`personal-auto-vmt-erosion-evidence`** — The first genuine crack in the premium pool: Progressive's September print showed personal-auto PIF **fell 27,000 sequentially** (28.195M), the first monthly decline since Nov 2023, with YoY growth decelerating every month since the 21.9% peak in March 2025. Q3 NPW growth slowed to +4% (September just +2%), stock fell 4.2% on the print. Not yet AV-attributable — Muse/AI-agent shopping and pricing are rival explanations — but the pool-growth-vs-adoption-curve crossover is now the metric to watch. 2. **`waymo-swiss-re-safety-claims-data`** — Waymo's 270M-mile safety update (82% fewer injury crashes, 95% fewer serious) is the actuarial core of the liability shift; three more cities opened to all, Toronto announced for 2028, 1M rides/week targeted by year-end — offset by a fresh NHTSA school-bus probe in Atlanta. 3. **`tesla-insurance-self-underwriting-liability-shift`** — Cybercab fleet grew **8x to 300+ in 30 days** (Texas registrations: 319 Cybercabs + 420 Model Y robotaxis), Dallas next; NHTSA school-zone petition review stacks on the Oct 30 Special Order deadline; the Washington filing shows Tesla pricing FSD miles preferentially while excluding gig liability entirely. 4. **`lemonade-autonomous-car-insurance-tesla-fsd`** — Lemonade extended per-autonomy pricing to HW3 Teslas (30% FSD-mile discount in AZ/CO/TN, "more coming soon"). **Watch bookkeeping:** Recorded events on the Q3-earnings watch (Progressive read done; Allstate/Kemper Nov 4, GEICO early Nov) and the Lemonade rollout watch. Left the Tesla-Washington and NHTSA Special Order watches open (neither has fired). Added three new watches: Tesla Q3 earnings (Oct 21), the NHTSA Waymo probe outcome, and Waymo crossing 1M weekly rides.

The research cycle is closed. Key takeaway for the thesis: this cycle produced the topic's first two-sided evidence — the adoption curve accelerated sharply (Tesla Cybercabs 8x in a month, Waymo opening cities and targeting 1M rides/week with 82%-fewer-injury actuarial data) while the personal-auto pool showed its first sequential policy decline since November 2023 at the largest writer, Progressive. Neither the erosion attribution nor the liability-shift timing is settled yet, but the next three catalysts are dated and now on the watch list: Tesla's Q3 call (Oct 21), the NHTSA Cybercab sworn answers (Oct 30), and Allstate/Kemper/GEICO earnings (Nov 4–early Nov).