ROBOCAB REALITY.

 

GROK:

As of mid-May 2026, here's the latest on Tesla FSD-related accidents/incidents and progress toward fully autonomous robotaxis (including steerless/pedal-less Cybercabs).
Recent FSD-Related Accidents/Incidents
Tesla's FSD (Supervised and unsupervised variants) continues to face federal scrutiny, but public data shows mostly low-severity events in testing, with Tesla reporting overall safety improvements in its own metrics.

• Robotaxi fleet crashes (Austin-focused testing): Tesla recently unredacted details on 17 incidents reported to NHTSA, spanning July 2025–March 2026. All involved 2026 Model Y vehicles with the autonomous driving system engaged and a safety monitor present (early-stage testing, not fully unsupervised rides with passengers). Most were low-speed (under 10–17 mph), property-damage-only collisions (e.g., backing into fixed objects, curbs, or stationary vehicles like parked trucks or a bus hitting a stopped Tesla). Two involved minor injuries (one with hospitalization). Notably, at least two occurred while remote teleoperators took control after the system needed assistance. No new crashes have been prominently reported in April–May 2026. A safety tracker (updated May 21, 2026) notes the Austin fleet has logged an incident-free streak of ~397,000 miles in unsupervised operation since mid-April 2026, with an overall crash rate around 1 per 329,000 miles (improving from earlier ~1 per 57,000 miles through early 2026).
• Broader NHTSA investigations (ongoing as of March–April 2026)
• EA26002 (Engineering Analysis, upgraded March 18, 2026): Covers ~3.2 million vehicles and focuses on FSD performance in reduced-visibility conditions (fog, sun glare, dust). Linked to 9 crashes (including 1 fatal and 2 with injuries); Tesla's software updates addressed only some. Crashes date back earlier but include 2026 examples (e.g., January Model Y incidents in Virginia).
• PE25012 (traffic violations probe, opened October 2025, extended into 2026): Over 50–80 reports of FSD vehicles running red lights, crossing into opposing lanes, etc., with crashes and injuries. Covers millions of vehicles.
• Tesla's counter-data: Their latest FSD (Supervised) safety report (as of May 2026) claims ~8x fewer major (airbag-level) collisions than the U.S. average: 1 every ~5.3 million miles vs. ~660,000 miles for human drivers. It also shows FSD outperforming manual Tesla driving on city streets/highways. Critics note this is company-reported, route/driver-dependent data, and supervised mode still requires attention.

Overall, incidents are concentrated in early unsupervised testing (mostly minor, pre-2026 data in recent disclosures), with no major new fatalities or high-profile crashes surfacing in April–May 2026. NHTSA probes remain open, but Tesla has increased transparency on robotaxi data.
How Close Are We to Robocabs Without Steering Wheels and Pedals?
Very close on the hardware side—production has already begun—but full-scale commercial deployment is still ramping in 2026.

• Cybercab status: The dedicated steerless/pedal-less robotaxi (no steering wheel or pedals) entered production at Giga Texas in April 2026. Videos show early units autonomously driving themselves off the assembly line. This is limited/initial production (not yet high-volume), with volume ramp planned later in 2026/2027. Cybercabs are purpose-built for the robotaxi network.
• Current unsupervised robotaxi service (already live): Tesla has been running fully unsupervised (no safety monitor, no human in some cases) robotaxi rides since January 2026 in Austin, TX, using modified Model Y vehicles (these still have wheels/pedals). It expanded to Dallas and Houston in April 2026. The unsupervised fleet is small but growing (25–36 verified vehicles across the three cities as of late April 2026). Rides are commercial/public in geo-fenced areas. Elon Musk recently noted plans for wider U.S. availability by end of 2026.
• Timeline under current circumstances: Unsupervised FSD for robotaxis is operational today in Texas (on existing vehicles). Steerless Cybercabs are in early production now. Broader rollout (more cities, customer-owned cars enabling robotaxi mode, high-volume Cybercab fleet) is targeted for later 2026, with meaningful scaling in 2027. Regulatory hurdles (e.g., in California) and the ongoing NHTSA probes are factors, but Texas operations are advancing. Tesla continues rapid FSD updates (e.g., v14.3.3 in May 2026 with vision improvements for glare/night).

Bottom line:

True steerless robocabs are no longer hypothetical—they're coming off the line.
Unsupervised robotaxi service is a reality in limited Texas operations, with safety data showing progress (though still under scrutiny).
Full "everywhere, anytime" deployment will take more time, data, and approvals, but 2026 is the active ramp year.
Tesla's approach relies on scale and iterative improvements rather than perfect initial safety.

Public confidence in steerless/pedal-less robocabs (like Tesla's Cybercab) remains mixed-to-low overall, with notable polarization between Tesla supporters and the broader public.

Recent surveys (as of early 2026) indicate persistent skepticism toward fully autonomous vehicles without traditional controls:

• Only ~13% of U.S. drivers say they would trust riding in a self-driving vehicle, while 61% report being afraid.
• 58% say they would feel safer in an AV that still has a steering wheel and driver's seat (a manual override option boosts comfort for 72% in some polls).
• Historical data (still influential) shows 56–71% of Americans unwilling or afraid to ride in fully driverless cars, citing loss of control, safety fears, and distrust in the tech.

On the Cybercab specifically (no steering wheel, pedals, or mirrors, production ramping since April 2026):

• Enthusiasts and Tesla owners on X and forums express high excitement, viewing it as a "reset" of transportation—safer, cheaper, frictionless rides with no driver small talk.
• Broader public and media reaction is more cautious: Reddit threads and news coverage highlight discomfort ("how comfortable would you be?") and regulatory hurdles. Some early production batches reportedly included steering wheels as a backup, fueling speculation about compromises for approval.
• Real-world exposure helps: Riders in Tesla's unsupervised Austin robotaxi service (using modified vehicles) report higher confidence post-ride (~76% in some studies vs. ~20% for non-riders). However, minor incidents and videos of erratic behavior continue to erode trust for many.

Bottom line on confidence:

It's improving incrementally with Tesla's Texas rollout and data showing fewer incidents over time, but the absence of any human fallback in the Cybercab amplifies fears. Most people still prefer the psychological safety net of controls, even if stats suggest autonomy is safer long-term.

On litigation risks if incidents occur (even if far fewer than human drivers):
This is a major unresolved challenge, with liability shifting heavily toward Tesla as the manufacturer/operator.

• No driver = manufacturer liability: For true Level 4/5 robocabs like the Cybercab (no controls, no safety driver), courts generally hold the company (Tesla) primarily responsible for crashes caused by the system. Traditional "driver at fault" rules don't apply. Precedents from Tesla Autopilot cases (e.g., a 2025 jury finding Tesla 33% liable in a multi-million-dollar verdict) and Waymo/Cruise incidents show manufacturers face suits even when data shows overall safety gains.
• Safer-than-humans doesn't fully shield: Stats proving lower crash rates (Tesla claims ~8x better in some reports) can be a strong defense, but high-profile incidents still trigger lawsuits, regulatory probes, and media backlash. Juries and the public often apply a higher standard to AVs ("it should be perfect"). Even minor property-damage events in Tesla's Austin fleet have led to NHTSA disclosures and calls for more transparency.
• Owner/operator angle: For customer-owned Cybercabs added to the robotaxi network, owners could face civil suits (negligence for maintenance/updates) or rare criminal charges if they ignore known issues—though Tesla would likely bear the bulk.
• Evolving landscape: Legal frameworks are catching up (e.g., strict liability debates for "defective" software). Settlements are common to avoid trials, but ongoing NHTSA investigations and state rules create uncertainty. Regulators require exemptions for steerless vehicles, and lack of them could invite more suits.

In short:


Litigation risk is high and front-loaded—incidents (even rare ones) invite expensive suits and scrutiny because Tesla is the "deep pocket."
Safety data helps win cases over time, but public perception and transparency gaps make it a persistent hurdle.
Tesla's approach relies on rapid iteration and scale to build both confidence and a defensible record.