Tesla Launched Cybercab—Then Washington Opened a Safety Audit

Tesla has initiated commercial Cybercab service in Austin, but faces an NHTSA audit over its lack of conventional driving controls. This indicates potential regulatory hurdles that could delay or prevent expansion to additional metro areas without compliance changes. The evidence does not confirm that Tesla will operate in at least three metro areas by March 31, 2027, without adding traditional controls or obtaining a model-specific federal exemption.

I select SCALES_AS_BUILT, though it is a close call against STAYS_A_PILOT. The decisive structural fact is that Tesla does not need to open new markets to reach three metros: Austin, Dallas, and Houston already carry paid unsupervised Tesla robotaxi service (Dallas and Houston launched unsupervised on April 18, 2026), and Florida metros followed. The Cybercab is already registered with the Texas DMV for commercial operation (~45 units at the September 3, 2026 Austin launch), with roughly 200-250 units observed staged at Giga Texas. Substituting Cybercabs into geofences, depots, and remote-ops infrastructure Tesla already runs is a far lower bar than the multi-city greenfield expansions Musk has repeatedly missed. Second, an NHTSA Audit Query is an information-gathering instrument, not a stop-work order; AQ26002 (opened September 3, 2026, covering ~1,000 vehicles) does not by itself bar continued paid service. The closest precedent is Zoox: it self-certified in July 2022, NHTSA issued a Special Order in September 2022, opened AQ23001 on March 3, 2023, and only closed it on August 4, 2025 — about 2.5 years from AQ opening — with Zoox operating throughout. A comparable process concluding inside the ~7 months to March 31, 2027 would be historically fast. Third, the federal posture is unusually permissive: NHTSA published NPRMs on March 16, 2026 excepting no-manual-control ADS vehicles from FMVSS 102, 103, and 104, proposed on June 26, 2026 to modernize FMVSS 135 for ADS-equipped vehicles, withdrew AV STEP, and issued AV Framework interim guidance and a Part 555 interim final rule on July 31, 2026. Critically, if those generic amendments are finalized they are rules of general applicability, not a 'model-specific federal exemption,' so they would not trigger the compliance-change branch as this question defines it. Fourth, Tesla has a concrete disincentive to seek Part 555 relief: the statutory cap of 2,500 vehicles per 12 months is roughly the fleet Tesla expects to field within a year, and Tesla has publicly framed self-certification as the path that avoids that cap. The main countervailing force is Tesla's own execution record — missed robotaxi guidance across multiple earnings calls, and an unsupervised fleet that was only ~20-30 active vehicles until a burst to ~200 immediately before the Cybercab event — plus the real possibility that NHTSA moves faster here than it did with Zoox, since it opened this AQ within one day rather than eight months. That risk is why my confidence is only medium, but three metros of nominal paid Cybercab service is a low, narrative-valuable milestone that Tesla has strong incentive and adequate infrastructure to hit before March 31, 2027.

Verified facts show that paid Cybercab service currently exists only in limited parts of Austin, but Tesla already operates its Model Y Robotaxi network in Dallas and Houston, has begun Cybercab production, and has registered Cybercabs under its Texas authorization. I therefore select SCALES_AS_BUILT: Tesla is more likely than not to add paid Cybercab rides in those two established metros before NHTSA completes its audit or compels a pause, redesign, or model-specific exemption. This is a narrow forecast about reaching three metros, not about achieving large-scale or legally secure deployment.

Verified as of the Sept 5, 2026 cutoff: Tesla began paid Cybercab rides in limited parts of Austin on Sept 3, 2026, with roughly 45 Cybercabs registered in Texas, no steering wheel or pedals, and no model-specific federal exemption; NHTSA opened an audit of Tesla's FMVSS self-certification on Sept 4. I treat the audit as an information request rather than a stop order: under the same Trump DOT, NHTSA proposed removing the brake-pedal requirement in June 2026 and approved Zoox for 5,000 steering-wheel-free robotaxis in August 2026, so a rapid pre-expansion shutdown is less likely than a slow compliance review. Texas permits driverless operation statewide and Tesla already runs its broader unsupervised robotaxi network in Dallas and Houston, making Cybercab expansion to a third metro operationally easy. Production is small and the launch was muted, which keeps confidence medium, but the most probable path is that Tesla reaches Austin plus two more Texas metros as-built before any qualifying pause, redesign, or exemption. I select SCALES_AS_BUILT.

Tesla launched commercial Cybercab rides in Austin on September 3, 2026, relying on self-certification under Federal Motor Vehicle Safety Standards (FMVSS) despite omitting mandatory physical controls like steering wheels, pedals, and side mirrors. On September 3–4, NHTSA opened Audit Query AQ26002 into Tesla's self-certification. Under existing FMVSS rules and regulatory precedent set during NHTSA's audit of Amazon's Zoox, automakers cannot self-certify compliance by declaring explicit manual control rules inapplicable without a Part 555 exemption or traditional controls. Because Cybercab currently operates paid rides in only one metro area (Austin), regulatory enforcement, an administrative pause, filing/obtaining a Part 555 exemption, or adding physical controls before scaling to a third metro area will cause a compliance change to occur first.

As of 5 September 2026, Tesla is offering paid, steering-wheel-free Cybercab rides only in Austin, with about 45 Texas registrations, delayed volume production, and an open NHTSA audit of its FMVSS self-certification. The March 31, 2027 cutoff is roughly seven months away. Tesla can leave the Austin pilot running without adding conventional controls or taking a model-specific exemption, while growing Model Y robotaxis instead. That makes remaining below three paid Cybercab metros the most likely path. Multi-city Texas expansion without a compliance change is plausible but is constrained by fleet size, the audit, and Tesla’s own gradual-expansion notice. A Zoox-style forced exemption or a service pause is possible but less likely in this short window, because NHTSA has not grounded the fleet and is simultaneously rewriting ADS standards rather than immediately imposing a 2,500-vehicle cap.

Tesla launched paid Cybercab rides in Austin on 2026-09-03 and NHTSA opened Audit Query AQ26002 the same day over Tesla's FMVSS self-certification of a vehicle with no steering wheel or pedals. Three factors favor Tesla reaching paid Cybercab service in at least three U.S. metros before any qualifying compliance change: (1) Tesla deliberately self-certified to avoid the 2,500-unit/yr exemption cap and is mass-producing and pre-staging Cybercabs in multiple cities; (2) it already operates robotaxi infrastructure and logs unsupervised trips across Texas and Florida metros (Austin, Dallas, Houston, Miami, Orlando, Tampa) and holds Texas authorization for 420 robotaxis, so adding paid Cybercab rides in two more metros within seven months is operationally plausible; (3) the Trump-administration NHTSA is moving to legalize ADS-dedicated vehicles generally (FMVSS 135 brake-pedal NPRM June 2026, prior steering-wheel rule, Administrator Morrison publicly endorsing removal of manual-control requirements), which would validate the as-built design without constituting a model-specific exemption, and NHTSA audits typically take many months, well past March 2027, before any enforcement. The main risks to this view are Tesla's well-documented history of slower-than-promised robotaxi expansion and the possibility the audit escalates into a recall or forced pause. On balance I select SCALES_AS_BUILT with low confidence.

As of September 5, 2026, Tesla has launched paid Cybercab rides in Austin, Texas, marking its first metro area. The National Highway Traffic Safety Administration (NHTSA) immediately opened an audit into Tesla’s self-certification of the Cybercab, specifically questioning how Tesla determined that the vehicle—lacking traditional controls like a steering wheel, pedals, or mirrors—complies with Federal Motor Vehicle Safety Standards (FMVSS). Tesla has not obtained a model-specific federal exemption, nor has it added traditional controls to the Cybercab. Tesla’s strategy hinges on self-certification, arguing that the Cybercab complies with existing FMVSS without needing a special exemption, which would cap production at 2,500 vehicles annually. However, the NHTSA audit is ongoing, and no resolution or enforcement action has been announced. Tesla has not publicly named the next cities for Cybercab expansion, and all current paid rides are confined to Austin. While Tesla has secured regulatory approvals in Nevada and other states for its Robotaxi network (which uses Model Y vehicles), the Cybercab’s unique design and lack of traditional controls present a novel regulatory challenge. The key question is whether Tesla can expand paid Cybercab service to at least two additional U.S. metro areas by March 31, 2027, without first adding traditional controls or obtaining a model-specific federal exemption. The NHTSA audit’s outcome is uncertain, and Tesla’s self-certification could be invalidated, leading to a compliance change. Alternatively, Tesla may proceed with expansion under its current interpretation, but the regulatory risk is high. The lack of public announcements about imminent expansion beyond Austin, combined with the NHTSA’s scrutiny, suggests that Tesla is unlikely to reach three paid-service metros without a compliance change or exemption.

SCALES_AS_BUILT is my most likely outcome, but with low confidence. Tesla has already begun commercial Cybercab deployment in Austin while relying on self-certification rather than a model-specific exemption, and it says it intends to expand to additional vehicles and locations. NHTSA’s September 4 Audit Query is serious, but it is an investigation rather than an announced stop order; absent an interim enforcement action, seven months is enough for Tesla to extend the existing Robotaxi operating infrastructure into at least two more metros. The main risk is that NHTSA concludes Tesla’s treatment of control-related FMVSS requirements was invalid and forces a pause, redesign, or exemption before expansion.

Tesla launched paid Cybercab service in Austin on September 3, 2026, without traditional controls or a model-specific federal exemption, relying instead on self-certification under FMVSS [1][19]. However, NHTSA initiated an audit of this self-certification the following day, questioning the legal basis for omitting mandatory safety equipment like steering wheels and pedals [19][22]. While Tesla aims to expand to other metros, the active regulatory scrutiny creates a significant risk of enforcement action, mandated pauses, or required compliance changes (such as obtaining an exemption or adding controls) before further expansion. Given the precedent of regulatory hurdles for autonomous vehicles and the specific challenge to Tesla's novel legal interpretation, it is more likely that Tesla will face a compliance barrier or pause before successfully scaling to three distinct metro areas by March 31, 2027, than that it will scale unimpeded. The 'Stays a Pilot' outcome is also plausible if expansion is slow, but the regulatory trigger makes 'Compliance Change Comes First' the most distinct risk-adjusted forecast for the specific condition of 'without... exemption'. However, looking at the options, 'Stays a Pilot' implies no change but failure to scale. 'Compliance Change' implies a regulatory stop. Given NHTSA's audit is not yet a ban, but a review, and Tesla has state-level approvals (e.g., Nevada, Texas), expansion might continue during the audit. Yet, the question asks if it will operate in 3 metros *without* first adding controls or getting an exemption. If NHTSA forces an exemption or redesign, that's Compliance Change. If they don't, but Tesla is slow, it's Stays a Pilot. Current reports indicate only Austin has Cybercabs, with Model Ys elsewhere. Expansion to 3 metros with *Cybercab* specifically in <7 months is aggressive. The regulatory cloud makes 'Compliance Change Comes First' a strong candidate if NHTSA acts decisively. However, NHTSA audits can take months. If Tesla expands to Phoenix or Las Vegas (where Zoox operates and regulations are friendlier) before a ruling, it could hit 3. But the prompt specifies 'without... obtaining a model-specific federal exemption'. Tesla explicitly avoided this path. If NHTSA rejects the self-certification, Tesla *must* get an exemption or redesign. This is a binary regulatory gate. Given the novelty and NHTSA's immediate action, the probability of a forced compliance change before full 3-metro scale is high.

Following the September 3, 2026 commercial Cybercab launch in Austin, NHTSA's immediate opening of an audit into Tesla's self-certification for a vehicle without conventional driving controls creates significant near-term regulatory risk that will prevent the unmodified rollout to three U.S. metro areas before March 31, 2027, without first requiring a compliance adjustment.

Tesla self-certified the Cybercab, a vehicle with no steering wheel, pedals, or mirrors, as FMVSS-compliant and began paid Austin service on 2026-09-03. NHTSA opened Audit Query AQ26002 that same day to examine whether Tesla's determination that certain FMVSS are inapplicable is valid. The only clear precedent, Zoox, was required to obtain a model-specific NHTSA exemption (granted 2026-07-31) for its no-controls robotaxi rather than rely on self-certification. With only about 1,000 Cybercabs produced and a seven-month window to 2027-03-31, a compliance event (exemption pursuit, pause, or added controls) preceding organic three-metro scale is more probable than Tesla reaching three paid metros as-built, though both timelines remain uncertain.

Cybercab paid service launched in Austin on September 3, 2026 with no conventional driving controls, and NHTSA opened an audit of Tesla's self-certification just one day later, signaling elevated federal scrutiny of the exact legal basis (FMVSS self-certification for a controls-free vehicle) that any multi-metro expansion would rest on. Precedent cuts against rapid scaling: Tesla's own Austin robotaxi rollout in mid-2025 began with a narrow geofence, invited riders, and onboard safety monitors, and only gradually loosened restrictions; Waymo's multi-metro driverless expansion took years; and state-level permitting for driverless vehicles without manual controls (especially California's DMV/CPUC regime) is slow and politically fraught. Reaching fully paid, controls-free Cybercab service in three U.S. metros within roughly seven months of the first launch would be an unprecedented pace for any AV operator. Tesla clearly intends to scale fast and Cybercab production is ramping, so a third metro is plausible, but the most likely outcome through March 31, 2027 is that the footprint stays at one or two metros, with the NHTSA audit, state permits, and Tesla's own cautious pilot-then-expand pattern acting as friction rather than a hard stop. A formal regulatory pause, hardware redesign to add controls, or a model-specific FMVSS exemption before a third metro is the least likely path, since none of the parties appears to want that outcome and self-certification lets Tesla proceed unless NHTSA forces a recall or injunction.

As of September 5, 2026, Tesla's Cybercab operates paid service only in Austin, Texas, with NHTSA having opened an audit of its self-certification the day after launch. The vehicle lacks conventional controls, and federal safety standards largely assume human-operated vehicles. Tesla has only 45 Cybercabs registered in Texas, with no evidence of paid service in Dallas, Houston, or other metros as of the cutoff. The NHTSA audit creates significant regulatory uncertainty that will likely require Tesla to address compliance before expanding to a third metro by March 31, 2027.