← back to scans
Will Tesla robotaxi be offered to the general public in California before Jan 1, 2027?
5%
Agent
8%
Market Price
-2.6%
Edge
79%
Confidence
15,881
Volume
3.0c
Spread
116
Days to resolution
4
Markets in event
final rationale
Mean of 5 draws (0.03, 0.03, 0.08, 0.08, 0.05). (0) The condition is not met. The CPUC (March 2026), Tesla's own filings, and trackers through October 4, 2026 all confirm the California service runs with a human driver under a limo-style TCP permit. Tesla holds only a DMV testing-with-driver permit and has not applied for driverless testing, deployment, or CPUC AV passenger permits. (a) Under three months remain before January 1, 2027. (b) Status quo is NO: a chauffeured service fails the driverless criterion. (c) NO scenario: Tesla keeps operating with safety drivers while R.25-08-013 stays undecided, which is by far the likeliest path. (d) YES scenario: Tesla files immediately and DMV/CPUC fast-track driverless and deployment approval, then a public fared launch follows before year-end. That sequence took Waymo and Cruise years and would require roughly 50k supervised test miles Tesla lacks. The Kalshi anchor is about 8% (bid 4%, ask 7%), and Polymarket's similar but possibly looser-defined market sits at 13%. Given the verified absence of any application, I price below the market at about 3%, leaving some room for unverified surprises such as an Oct 21 earnings-call announcement or a regulatory shortcut.
price vs. prediction
market price
agent forecast at t+0
12 follow-ups
1. Planner Opus 5.5 · picks the research tools
Tools chosen:
kalshi_direct
resolution_source
kalshi_related
polymarket_related
article_search
Planner reasoning
Kalshi market price is the anchor. Related Kalshi robotaxi-area markets and Polymarket give cross-market signal. The key fact is whether Tesla holds a California DMV driverless deployment permit and a CPUC driverless passenger permit, which is the regulatory gate. No resolution URL is named, so I've given a source name. Tesla stock data adds minor context.
2. Research parallel tool calls + 4 web research loops (Sonnet 5) · read raw by the forecasters
at a glance
RESEARCH DIGEST (built in code: tool summaries and each report's bottom line; the forecasters read the full tool outputs and reports)
Tools
- kalshi_direct: ## This Market's Kalshi Data **California** (KXROBOTAXIAREA-26DEC-CA) - Current price (probability): 8.00% (live; bid 4% / ask 7%) - Last daily close (2026-09-22): 4.00% - 7-day price change: +2.00% - 30-day price change: -7.00% - Average daily volume: 239 contracts - Price range: 4.00% - 17.00% - Data points: 37 days
- resolution_source: already_resolved=no; matches_criteria=True; last_updated=2026-03-25. As of March 2026, CPUC officials confirm Tesla does not hold an AV permit and is operating under a Charter-Party Carrier (TCP) permit—the same type as a limousine company—with a human driver in the driver's seat (not a safety driver, but 'the driver'). Tesla's robotaxi rides in Texas and the Bay Area rely on in-car human drivers
- kalshi_related: 0 related markets / summaries. series KXROBOTAXIAREA: 0 markets (skipped 4 no-signal) | keyword 'Tesla robotaxi': no matches | keyword 'Tesla unsupervised': no matches
- polymarket_related: Polymarket search over 3 keyword(s), kept 3 matches. keyword 'Tesla robotaxi': 2 markets | keyword 'Tesla California': 1 markets | keyword 'Tesla unsupervised FSD': 0 markets
- article_search: Fetched 5 unique articles across 2 queries (lookback=120d). 'Tesla California DMV driverless deployment permit': 8 articles | 'Tesla CPUC driverless passenger permit application': 8 articles
Research loops, bottom lines
- current state: As of the most recent verifiable reporting (through early October 2026), Tesla's California "Robotaxi" service remains a human-driven, fare-charging chauffeur service under a TCP (limousine-type) permit; it has neither applied for nor received the CPUC/DMV autonomous-vehicle permits required to operate without a driver, and every tracked source confirms a safety driver is legally required in the Bay Area — meaning none of the "offered to" criteria (driverless operation) is currently met in Calif
- base rate: The reference class of "AV company reaching fared, driverless, general-public deployment in California" has exactly one precedent pair (Waymo, Cruise), and it took roughly 5 years from initial pilot framework to driverless-fared authorization, built on millions of supervised/driverless test miles and multiple sequential CPUC/DMV permits; Tesla has not entered that pipeline at all (zero recent CA test miles, no AV permit application, still just a TCP/limo permit), making a sub-3-month jump to ful
- yes case: The only plausible YES path is a sudden Tesla filing-and-approval sprint — leveraging its proven ability to flip to driverless fared service quickly once cleared (as in Austin, Sept 2026) and an active CPUC rulemaking it is already shaping — but as of Oct 7 2026 Tesla has filed no DMV driverless-permit or CPUC AV-passenger-service application, so YES requires an application, approval, and public launch all within under three months, a sequence with no precedent in the state's history.
- no case: Every data point through mid-October 2026 — Tesla's own CPUC filings, the regulator's public statements, the SFO permit terms, and Wikipedia's service-status table — confirms California remains the one major Tesla robotaxi market still legally requiring a human driver, with no AV permit application pending and a rulemaking (R.25-08-013) still open with no decision date; absent an extremely fast-tracked, historically unprecedented CPUC/DMV approval in under three months, the "offered to the gener
tools
| tool | status | time | items | summary |
|---|---|---|---|---|
| kalshi_direct | OK | 0.9s | 1 | ## This Market's Kalshi Data **California** (KXROBOTAXIAREA-26DEC-CA) - Current price (probability): 8.00% (live; bid 4% / ask 7%) - Last daily close (2026-09-22): 4.00% - 7-day price change: +2.00% - 30-day price change: -7.00% - Average daily volume: 239 contracts - Price range: 4.00% - 17.00% - |
| resolution_source | OK | 42.6s | 1 | already_resolved=no; matches_criteria=True; last_updated=2026-03-25. As of March 2026, CPUC officials confirm Tesla does not hold an AV permit and is operating under a Charter-Party Carrier (TCP) permit—the same type as a limousine company—with a human driver in the driver's seat (not a safety drive |
| kalshi_related | OK | 37.1s | 0 | 0 related markets / summaries. series KXROBOTAXIAREA: 0 markets (skipped 4 no-signal) | keyword 'Tesla robotaxi': no matches | keyword 'Tesla unsupervised': no matches |
| polymarket_related | OK | 0.4s | 3 | Polymarket search over 3 keyword(s), kept 3 matches. keyword 'Tesla robotaxi': 2 markets | keyword 'Tesla California': 1 markets | keyword 'Tesla unsupervised FSD': 0 markets |
| article_search | OK | 0.7s | 5 | Fetched 5 unique articles across 2 queries (lookback=120d). 'Tesla California DMV driverless deployment permit': 8 articles | 'Tesla CPUC driverless passenger permit application': 8 articles |
research loops
·
4/4 ok ·
18 searches ·
5 page reads
| dimension | status | time | searches | page reads | report |
|---|---|---|---|---|---|
| current state | OK | 32s | 4 | 2 | 5457 chars |
| base rate | OK | 28s | 4 | 2 | 4296 chars |
| yes case | SNIPPETS | 38s | 5 | 0 | 4957 chars · no page read; snippets only |
| no case | OK | 31s | 5 | 1 | 4832 chars |
report: current state
- [2025-07-29] Tesla received a CPUC transportation charter-party carrier (TCP) permit — a chauffeur/limousine-type license, not an autonomous-vehicle permit; CPUC stated the permit
"does not authorize them to provide rides" in autonomous vehicles, and does not allow Tesla to operate a ride-hailing service to the public
. https://www.newsweek.com/tesla-california-robotaxi-permit-driverless-cars-2047030
- [2025-07-31] Tesla launched a Bay Area "Robotaxi" ride-hailing service, but
a driver will currently sit behind the wheel due to the company lacking the proper permits to operate autonomous rides
; access was initially invite-only. https://abc7news.com/post/tesla-launches-robotaxi-service-bay-area-drivers-questions-surround-autonomous-permitting/17377400/
- [2026-03-24] CPUC Deputy Executive Director Pat Tsen stated on a podcast that Tesla's California service is legally a limo/chauffeur service, not an AV:
"What they have from us is essentially a charter party carrier permit. It is the same type of permit that a limousine company would get... So, in terms of our view of the person who is sitting in the driver's seat, that is a driver that is not a safety driver"
. https://gizmodo.com/california-regulator-says-teslas-robotaxis-are-more-like-a-limo-in-the-eyes-of-the-law-2000738512
- [2026-02-01] Tesla's own CPUC filing conceded its "Robotaxi" service
still relies on in-car human drivers and remote operators in both Austin and the Bay Area... explicitly acknowledged using an SAE Level 2 system
. https://electrek.co/2026/03/25/california-regulator-confirms-tesla-not-operating-autonomous-vehicle-service/
- [2026-05 (as of)] California DMV records show
Tesla Robotaxi LLC appears on the California DMV list of Autonomous Vehicle Testing Permit holders (with a driver), as of May 2026. The company has not obtained driverless testing or deployment permits
. https://notanfsdtracker.com/united-states-tesla-robotaxi-rollout-2026
- [2026-05/Q2 filing] Tesla's own SEC 8-K investor deck lists its planned robotaxi coverage table with California/SF Bay Area status marked
"Safety Driver"
while Texas, Arizona, Florida and Nevada metros are listed as unsupervised or ramping-unsupervised. https://www.sec.gov/Archives/edgar/data/1318605/000162828026003837/exhibit991.htm
- [2026-06/Q2] Tesla's Bay Area ride-hailing expanded operationally (e.g., to San Jose Airport) but remained supervised;
"Our Bay Area ride-hailing service began serving the San Jose Airport in October, with plans to expand to other major airports in the Bay Area upon receiving required permitting"
(permitting still pending). https://www.sec.gov/Archives/edgar/data/1318605/000162828026003837/exhibit991.htm
- [2026-09-24 (verified date)] Independent tracker confirms as of October 2026: Tesla Robotaxi operates in 7 metros, but is rated "Limited" not "Active" because
"since September 4, 2026, the purpose-built Cybercab is carrying public riders in a limited part of Austin... Tesla began removing safety monitors in Austin in January 2026 on a few vehicles... the Bay Area still requires a safety driver by law"
. https://thechargeport.com/robotaxi-tracker
- [2026-10-04 (last checked)] Same tracker lists California explicitly as
"San Francisco Bay Area, CA (safety driver — CA law)"
, distinct from fully unsupervised Texas/Florida metros, with no change reported in the most recent update. https://thechargeport.com/robotaxi-tracker
- [2026-09-21] San Jose Inside republished/updated coverage reiterating CPUC's position:
"Tesla has not received approval from the CPUC to offer autonomous passenger service to customers, paid or unpaid, with or without a driver (nor has the company applied)."
No indication an AV application has since been filed. https://www.sanjoseinside.com/business/teslas-bay-area-employees-only-robotaxi-rollout-in-bay-area-has-human-driver/
- [2026-03-27] Analysis concluded
"Tesla will not receive a CPUC autonomous vehicle permit for its Bay Area operation before the end of 2026. The company hasn't applied for one."
No subsequent reporting found that contradicts this. https://evxl.co/2026/03/27/tesla-cpuc-not-robotaxi-california-limo-permit/
- [2026-03] Comparative regulatory context:
"Tesla cannot operate without a driver in California until it earns the proper permits... Alongside Zoox and the recently-revived Cruise, [Waymo] is one of three companies with full California deployment permits"
— Tesla is not among them. https://www.automotiveworld.com/news/teslas-california-robotaxi-service-is-still-a-long-way-off/
BOTTOM LINE: As of the most recent verifiable reporting (through early October 2026), Tesla's California "Robotaxi" service remains a human-driven, fare-charging chauffeur service under a TCP (limousine-type) permit; it has neither applied for nor received the CPUC/DMV autonomous-vehicle permits required to operate without a driver, and every tracked source confirms a safety driver is legally required in the Bay Area — meaning none of the "offered to" criteria (driverless operation) is currently met in California, unlike Tesla's unsupervised operations in Texas/Florida.
NOT FOUND: No evidence of any new CPUC or DMV application by Tesla for driverless testing/deployment permits in California filed in 2026; no announced or scheduled date for a driverless Bay Area launch; no official CPUC decision on Tesla's "Robotaxi" trademark/marketing dispute as of the search date.
report: base rate
- [2026-02-27] Tesla
logged zero miles of autonomous test driving on California roads in 2025, and has not reported any autonomous test miles in the state since 2019, with just 562 miles total since 2016
.
Proposed DMV rules expected to be finalized this year would require 50,000 autonomous safety-driver miles before applying to test driverless; Tesla has not met that threshold
. (cbtnews.com)
- [2026-02-27] Reference class comparison:
Waymo logged more than 13 million autonomous testing miles between 2014 and 2023 and secured seven regulatory approvals before receiving permission to charge passengers for driverless rides
. (cbtnews.com)
- [2022-02-28] Historical base rate for the AV-industry pathway:
CPUC issued Drivered Deployment permits to Cruise and Waymo to allow passenger service with a safety driver present and fare collection, years after their initial pilot permits
. (cpuc.ca.gov)
- [2023-08-10]
After more than five years of implementation and regulatory processes, CPUC voted to approve resolutions allowing Cruise and Waymo to operate like taxis and charge passenger fares at any hour in San Francisco
, the first (and so far only) companies to clear the full "offered to general public, driverless, fared" bar matching this question's criteria. (gtlaw.com)
- [2026-03-24] CPUC's deputy executive director for transportation, Pat Tsen, stated on record: "Tesla is not operating an autonomous vehicle service... Tesla is a level two system and so they do not have a permit with the California Public Utilities Commission and my understanding is that they do not have a permit with the DMV either." (electrek.co)
- [2026-03-24] Tsen explained Tesla's only authorization is
"essentially a charter party carrier permit. It is the same type of permit that a limousine company would get from the CPUC to provide a limousine service. So, in terms of our view of the person who is sitting in the driver's seat, that is a driver that is not a safety driver"
. (gizmodo.com)
- [2026-02-13]
In a CPUC filing (Rulemaking 25-08-013), Tesla acknowledged in writing that its ride-hailing service in both Austin and the Bay Area relies on in-car human drivers and remotely located operators, running on an SAE Level 2 system
. (evxl.co)
- [2026-01-30]
Tesla itself said in a filing to the CPUC that it "does not currently possess the requisite permits to operate autonomous rideshare vehicles in California" and that all current rides are conducted with a safety driver in the driver's seat using Full Self-Driving (Supervised)
. (tesorb.com)
- [2026-01-01] Tesla's own shareholder update described Austin as "ramping unsupervised" but
listed the San Francisco Bay Area as "safety driver"
— i.e. by Tesla's own characterization, California is the laggard market, not the lead one. (tesorb.com)
- [2026-03-20]
The California DMV's official permit-holder page lists Tesla Robotaxi LLC as a permit holder for testing with a driver only, not among companies approved for driverless testing or deployment
. (tesorb.com)
- [2025-12-15 to 2026-01] Outside Texas, the only jurisdiction where Tesla has removed safety monitors at all is Austin (and limited expansions to Dallas/Houston), where
Tesla Robotaxi LLC holds a TNC permit and is operating limited robotaxi services with some vehicles running without safety drivers under state permits
, a much lighter-touch regulatory regime than California's.
BOTTOM LINE: The reference class of "AV company reaching fared, driverless, general-public deployment in California" has exactly one precedent pair (Waymo, Cruise), and it took roughly 5 years from initial pilot framework to driverless-fared authorization, built on millions of supervised/driverless test miles and multiple sequential CPUC/DMV permits; Tesla has not entered that pipeline at all (zero recent CA test miles, no AV permit application, still just a TCP/limo permit), making a sub-3-month jump to full compliant deployment by Jan 1, 2027 an extreme outlier versus all historical analogues.
NOT FOUND: No public prediction-market, pollster, or analyst base-rate estimate specifically modeling "Tesla driverless fared CA launch by date X" was found; no evidence Tesla has filed a CPUC AV Drivered/Driverless Pilot or Deployment application as of early October 2026.
report: yes case
YES means: Before Jan 1 2027, Tesla operates a fully driverless (no human in the driver's seat controlling the vehicle), fare-charging ride-hailing service open to the general public (no invite/employee restriction) somewhere in California, satisfying all three Kalshi criteria simultaneously.
- [2026-01-30] Tesla's own CPUC filing in the open Rulemaking 25-08-013 shows the company is an active, lobbying participant in the exact proceeding that will decide how "Level 2" ride-hailing like its Bay Area service is regulated — i.e., there is a live, Tesla-engaged regulatory track that could in theory produce a faster, lighter-touch authorization than the full AV pilot/deployment pathway Waymo/Cruise used. https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M598/K100/598100914.PDF
- [2025-11-17] In Reply Comments on the same docket, Tesla argued against CPUC treating Level 2 ride-hailing the same as true AVs, seeking continued light regulation of its existing TCP-permitted service — a path that, if adopted, would not require it to go through DMV's driverless-testing/deployment ladder at all. https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M587/K755/587755957.PDF
- [2026-09-04] Tesla's Cybercab began carrying public riders without safety monitors in part of Austin, demonstrating the company can and does execute a driverless, fared, public launch once it clears a given jurisdiction's regulatory bar — the Austin precedent is the template Tesla would replicate in California if permits were obtained. https://thechargeport.com/robotaxi-tracker
- [2026-08-14] Even as CPUC confirmed Tesla has no AV permit, the same month saw Waymo win expanded CPUC approval across 18 counties, showing CPUC is actively approving new driverless-fare authorizations in 2026 — the agency is not frozen, so a Tesla application, if filed, would not face a dormant process. https://electrek.co/2026/08/14/waymo-cpuc-approval-california-expansion-18-counties/
- [2026-05 SEC 8-K] Tesla stated its Bay Area service has "plans to expand to other major airports in the Bay Area upon receiving required permitting," indicating the company is actively pursuing additional CPUC authorizations rather than standing still. https://www.sec.gov/Archives/edgar/data/1318605/000162828026003837/exhibit991.htm
- [2025-10-22] On the Q3 2025 earnings call, Musk's team projected safety-driver removal "within 2.3 months" for other metros and signaled a pattern of starting with a 3-month supervised period before removing the driver in each new market — if Tesla applied for and received a CPUC/DMV AV permit soon, this self-described cadence could in principle compress a California driverless launch into Q4 2026/Q1 2027. https://www.forbes.com/sites/bradtempleton/2025/10/22/tesla-earnings-call-offers-self-driving-updates-similar-predictions/
- [2026-10-21 (scheduled)] Tesla's Q3 2026 earnings call, falling before the Jan 1 2027 deadline, is a concrete near-term event at which management could announce a filed CPUC/DMV application or an accelerated California timeline, given Musk's history of using earnings calls to announce new regulatory/launch milestones. https://www.teslaoracle.com/2026/10/04/tesla-tsla-announces-the-date-and-time-for-the-q3-2026-earnings-call-live-webcast-with-elon-musk-and-his-team/
- [2026-02-27] Reporting notes the DMV's 50,000-mile safety-driver-mile threshold for driverless-testing eligibility is still only "proposed," not finalized — meaning the exact bar Tesla must clear remains in flux and could in principle be set, interpreted, or waived in a way more favorable to Tesla before year-end. https://www.cbtnews.com/tesla-california-robotaxi-push-stalls/
- [2026-03] A DMV spokesperson told reporters the agency "is ready to assess applications," underscoring that the sole blocking factor is Tesla's own non-filing, not an agency backlog — a single Tesla filing decision (not a multi-year bureaucratic wait) is the only procedural step identified as missing. https://www.automotiveworld.com/news/teslas-california-robotaxi-service-is-still-a-long-way-off/
BOTTOM LINE: The only plausible YES path is a sudden Tesla filing-and-approval sprint — leveraging its proven ability to flip to driverless fared service quickly once cleared (as in Austin, Sept 2026) and an active CPUC rulemaking it is already shaping — but as of Oct 7 2026 Tesla has filed no DMV driverless-permit or CPUC AV-passenger-service application, so YES requires an application, approval, and public launch all within under three months, a sequence with no precedent in the state's history.
NOT FOUND: No evidence Tesla has filed (or announced intent to file) a DMV driverless-testing/deployment application or a CPUC AV Passenger Service Program application as of the search date; no outcome yet from CPUC Rulemaking 25-08-013 that would reclassify Tesla's Level 2 service; no confirmed agenda item on Tesla California robotaxi for the Oct 21, 2026 earnings call.
report: no case
YES means: Before Jan 1, 2027, Tesla operates in California a ride-hailing service that (a) is open to the general public via an app, (b) runs with no human driver/safety monitor in the driver's seat actively controlling the vehicle, and (c) charges fares — i.e., a genuine driverless, fared Tesla robotaxi, not the current chauffeured service.
- [2026-08-12] As recently as mid-August 2026, a new Tesla Cybercab spotted on San Francisco streets still showed a human behind the wheel, and the SF Chronicle noted Tesla's permit only lets it carry passengers with a human driver, even for its new SFO airport run launched the prior month. https://www.sfchronicle.com/sf/article/tesla-robotaxi-gold-22385407.php
- [2026-07-22] SFO airport itself confirmed Tesla's service there runs only under a "limo permit" with no autonomous operations authorized, meaning even Tesla's newest market expansion in California is explicitly human-driven, not unsupervised. https://www.kron4.com/news/technology-ai/tesla-robotaxi-service-now-goes-to-sfo/
- [2026-09-xx] Wikipedia's maintained service-area table (updated through September 2026) lists every Texas and Florida metro as "Full commercial service" (unsupervised) but tags California/Bay Area only as "Safety-driver service" with no launch date — the only active U.S. market still in that category. https://en.wikipedia.org/wiki/Tesla_Robotaxi
- [2026-01-30] Tesla's own CPUC prehearing filing in Rulemaking 25-08-013 (filed Jan 30, 2026) continued to argue Level-2 ADAS operators like itself shouldn't face the same reporting/consent rules as true AVs — a position inconsistent with imminently seeking driverless authorization. https://docs.cpuc.ca.gov/PublishedDocs/Efile/G000/M598/K101/598101354.PDF
- [2025-09-05] The governing CPUC rulemaking (R.25-08-013) was only opened August 28, 2025, as a successor to a prior proceeding that took from 2012 to 2023 (per colleagues' Waymo/Cruise base rate) to produce full driverless-fared authorization — this new rulemaking has not yet issued a final decision as of the search date, and no decision timeline before 2027 was found. https://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M578/K911/578911303.PDF
- [2026-02-13/19] Tesla's own CPUC filing conceded its service "still relies on both in-car human drivers and domestic remote operators to function," with remote operators required to hold driver's licenses and pass background/drug checks — the opposite of a driverless architecture. https://electrek.co/2026/02/19/tesla-admits-needs-drivers-and-remote-operators-argues-better-than-waymo/
- [2026-02-17/18] Separately, Tesla was simultaneously fighting a DMV false-advertising sanction over "Autopilot"/"Full Self-Driving" marketing, only narrowly avoiding a 30-day sales/manufacturing license suspension by dropping the "Autopilot" name — showing the regulatory relationship is adversarial, not one of imminent cooperation toward an AV deployment permit. https://goldrushcam.com/sierrasuntimes/index.php/news/local-news/75606-california-department-of-motor-vehicles-reports-tesla-takes-corrective-action-to-avoid-dmv-suspension
- [2026-09 (per lawsuit tracker, "as of September 2026")] Tesla is also still litigating the DMV's false-advertising ruling (suing to reverse it) and facing active federal class actions over FSD/ADAS marketing — ongoing legal friction that gives California regulators added reason for caution before granting Tesla a driverless AV permit. https://www.recordinglaw.com/us-laws/self-driving-car-laws/tesla-autopilot-fsd-lawsuits/
- [2026-04/2026 mechanism] By contrast, every actual Tesla driverless milestone this year (Austin unsupervised Jan 2026, Dallas/Houston April 2026, Miami/Orlando/Tampa July 2026, Cybercab Sept 2026) has occurred in lighter-touch states (Texas, Florida) with TNC-style permitting, not in California's multi-stage CPUC/DMV pilot-to-deployment pathway — the mechanism that unlocked those states does not exist in California's framework.
BOTTOM LINE: Every data point through mid-October 2026 — Tesla's own CPUC filings, the regulator's public statements, the SFO permit terms, and Wikipedia's service-status table — confirms California remains the one major Tesla robotaxi market still legally requiring a human driver, with no AV permit application pending and a rulemaking (R.25-08-013) still open with no decision date; absent an extremely fast-tracked, historically unprecedented CPUC/DMV approval in under three months, the "offered to the general public" driverless-fare bar is not met and the question resolves NO.
NOT FOUND: No evidence of a Tesla CPUC/DMV driverless-testing or deployment permit application filed as of the search date; no scheduled CPUC decision date in R.25-08-013 before Jan 1, 2027; no announced date for removing the Bay Area safety driver.
3. Forecasters 5 draws · Opus 5.5 + GPT-6 Astra + GPT-5.6 Sol · plain mean
claude-opus-5-5 · high effort
85% conf
| Yes | 3% |
| No | 97% |
(0) The condition is not met. The CPUC (March 2026), Tesla's own filings, and trackers through October 4, 2026 all confirm the California service runs with a human driver under a limo-style TCP permit. Tesla holds only a DMV testing-with-driver permit and has not applied for driverless testing, deployment, or CPUC AV passenger permits. (a) Under three months remain before January 1, 2027. (b) Status quo is NO: a chauffeured service fails the driverless criterion. (c) NO scenario: Tesla keeps operating with safety drivers while R.25-08-013 stays undecided, which is by far the likeliest path. (d) YES scenario: Tesla files immediately and DMV/CPUC fast-track driverless and deployment approval, then a public fared launch follows before year-end. That sequence took Waymo and Cruise years and would require roughly 50k supervised test miles Tesla lacks. The Kalshi anchor is about 8% (bid 4%, ask 7%), and Polymarket's similar but possibly looser-defined market sits at 13%. Given the verified absence of any application, I price below the market at about 3%, leaving some room for unverified surprises such as an Oct 21 earnings-call announcement or a regulatory shortcut.
claude-opus-5-5 · high effort
85% conf
| Yes | 3% |
| No | 97% |
(0) Not met: every source through early October 2026 (CPUC statements, Tesla's own CPUC filings and SEC deck, trackers updated Oct 4) shows the California service still runs with a human driver under a limo-type TCP permit. Tesla has not applied for DMV driverless or CPUC AV deployment permits. (a) Under three months remain before Jan 1, 2027. In that window Tesla would need to file, get DMV driverless testing and deployment permits plus CPUC driverless pilot and deployment approval, and launch a public, fared, driverless service. (b) Status quo is No: the chauffeured service fails the no-driver criterion. (c) No scenario: Tesla keeps its supervised Bay Area service while expanding driverless in Texas and Florida, and the CPUC rulemaking R.25-08-013 stays undecided. (d) Yes scenario: an unprecedented fast-track in which Tesla files after the Oct 21 earnings call, regulators approve within weeks, and a driverless fared launch follows before year-end. Waymo and Cruise took years of permits and millions of test miles to reach this point, while Tesla logged zero California AV test miles in 2025. The Kalshi price is about 8% (bid 4 / ask 7, last close 4%), and Polymarket's similar but less strictly worded market sits at 13%. Given the absence of any application, I price below the market, at about 3%.
gpt-6-astra · high effort
72% conf
| Yes | 8% |
| No | 92% |
I anchor at Kalshi's reported 8% YES price, retaining it because the research supports the low consensus probability without establishing a clearly unpriced catalyst. The supplied March 25 resolution-source read documents driver-operated service rather than a qualifying launch, and the October 4 tracker still labels California operations safety-driver service, so YES is neither already met nor impossible. Approximately 86 days remain until January 1, 2027; the January 31 trading close does not extend that deadline. If nothing changes, NO wins because even publicly available, paid rides fail the rules while an onboard safety driver or operator can intervene. A straightforward NO scenario is continued supervised operation without driverless authorization, consistent with July SFO permit reporting and the historically multistage DMV/CPUC approval process faced by Waymo and Cruise. A plausible YES scenario is expedited authorization followed by unrestricted, fare-charging public rides somewhere in California before year-end, with neither an onboard safety operator nor continuous remote human control; CNBC's October 3 reporting on Austin driverless Cybercabs supports operational feasibility. Tesla's Level-2 lobbying does not itself authorize driverless service, while the proposed 50,000-mile testing threshold should not be treated as an established binding barrier. I explicitly allow for unreported permitting progress and documentation lag because the resolution-source read is six months stale, some September coverage republishes older claims, and the research did not directly verify a fresh official DMV/CPUC permit register.
gpt-6-astra · high effort
69% conf
| Yes | 8% |
| No | 92% |
Kalshi's reported 8% YES is my anchor, and I retain it because the research supplies no clearly unpriced development warranting a change. No source documents a qualifying California launch: the March 25 resolution-source read reports a chauffeur operation, and the October 4 tracker still lists a safety driver, so YES is not already met but remains possible. There are 86 days until January 1, 2027; January 31 is only the trading close. The status quo resolves No, as would continued driver supervision or a driverless pilot restricted to invitees through year-end. A YES scenario is accelerated DMV/CPUC approval followed by even a small, paid, publicly accessible California service with no onboard safety operator and no continuous remote driving; CNBC's October 3 reporting on Austin Cybercab operations provides a relevant operational precedent. California's multistage approvals and the lengthy Waymo/Cruise histories favor NO, while lobbying over Level 2 services does not itself authorize driverless deployment. Meaningful uncertainty remains because the official-backed resolution read is six months old and the claimed absence of applications lacks a fresh regulator-registry check; missing updates could change the outlook, but resolution still requires documented qualifying operations rather than branding or announcements. Polymarket's 13% is context only because matching settlement criteria were not verified.
gpt-5.6-sol · high effort
86% conf
| Yes | 5% |
| No | 95% |
(0) The condition has not been met: the October 4 tracker still lists the Bay Area service as requiring a safety driver, while September 21 reporting says Tesla has neither received nor applied for CPUC autonomous-passenger approval. (a) About 86 days remain until the before-January-1, 2027 deadline, during which Tesla would need to apply for and obtain the necessary DMV and CPUC approvals and then launch a fare-charging, non-invite-only driverless service. (b) If nothing changes, the existing TCP-permitted, human-driven California service fails the no-driver requirement and resolves No. Kalshi's consensus anchor is 8% Yes, while Polymarket's less explicitly defined California launch market is 13%, providing some independent support for a nonzero chance. I move below the Kalshi anchor because the post-September evidence still shows no application, Tesla's own filings characterize California as safety-driver service, and California precedents required years and extensive testing rather than a sub-three-month approval sprint. (c) The straightforward No scenario is that Tesla continues its supervised Bay Area service while regulatory applications or rulemaking extend into 2027. (d) Yes requires an abrupt filing and expedited approval, followed by a public paid launch without an onboard driver—conceivably aided by Tesla's Austin driverless experience and active CPUC lobbying, but historically unprecedented on this timeline.
4. Final plain mean of 5 draws
5%
(79% confidence)
Yes: 5%
No: 95%
Mean of 5 draws (0.03, 0.03, 0.08, 0.08, 0.05). (0) The condition is not met. The CPUC (March 2026), Tesla's own filings, and trackers through October 4, 2026 all confirm the California service runs with a human driver under a limo-style TCP permit. Tesla holds only a DMV testing-with-driver permit and has not applied for driverless testing, deployment, or CPUC AV passenger permits. (a) Under three months remain before January 1, 2027. (b) Status quo is NO: a chauffeured service fails the driverless criterion. (c) NO scenario: Tesla keeps operating with safety drivers while R.25-08-013 stays undecided, which is by far the likeliest path. (d) YES scenario: Tesla files immediately and DMV/CPUC fast-track driverless and deployment approval, then a public fared launch follows before year-end. That sequence took Waymo and Cruise years and would require roughly 50k supervised test miles Tesla lacks. The Kalshi anchor is about 8% (bid 4%, ask 7%), and Polymarket's similar but possibly looser-defined market sits at 13%. Given the verified absence of any application, I price below the market at about 3%, leaving some room for unverified surprises such as an Oct 21 earnings-call announcement or a regulatory shortcut.
Pipeline Timing
Total pipeline time: 117.7s
Per-tool research timings shown in the Research section above.