Tesla Autopilot and FSD Crashes Reach Unprecedented Levels
Rising Incidents: A Critical Look at Tesla's Driver-Assist Systems
Unprecedented Surge in Reported Incidents
Tesla has documented a record-breaking 207 accidents linked to its driver-assistance technologies in a single month. This substantial number, derived from Tesla's submissions to the National Highway Traffic Safety Administration (NHTSA), represents the highest monthly count ever recorded. The reports are mandated under NHTSA's Standing General Order, which requires manufacturers to report incidents where advanced driver-assistance systems (ADAS) were active shortly before impact and met specific severity criteria, such as airbag deployment, vehicle towing, injuries, fatalities, or striking pedestrians or cyclists.
Analyzing the Escalating Crash Data
By integrating NHTSA's most recent ADAS export with historical data dating back to 2019, and after removing duplicate entries, a clear pattern emerges concerning Tesla's Level 2 systems, specifically Autopilot and FSD used in customer vehicles. The data unequivocally shows that Tesla recorded 207 crashes in May 2026, marking an all-time monthly high. This single month's total exceeds the cumulative number of incidents reported for the entire year of 2021, which stood at 157. Currently, Tesla accounts for 3,763 unique driver-assist crashes, representing approximately 85% of all ADAS reports across the automotive industry. This trend indicates an accelerating, rather than stabilizing, increase in incidents.
Yearly Trends and Accelerated Growth
The year 2025 was the first year Tesla's driver-assist incidents surpassed 1,000, reaching 1,043. For 2026, the company has already reported 826 crashes within the first six months, putting it on track to potentially double the previous year's total. When comparing the same January-to-June period across years to ensure an equitable comparison, the figures demonstrate a sharp increase: 180, 261, 269, 476, and now 826. This represents a 73% year-over-year surge from an already elevated baseline.
The Challenge of Assessing Per-Mile Safety
An important consideration is that the raw number of crashes naturally increases with a larger fleet size and more miles driven. Tesla has expanded its vehicle sales and FSD subscriptions, leading to its systems being used more extensively on roads. While increased mileage might lead to more incidents, it does not necessarily imply a decrease in safety per mile. However, the exact per-mile crash rate remains opaque to external analysis because Tesla does not provide independently verifiable data on miles driven by Autopilot and FSD. The company's quarterly "safety report" has faced criticism for its potentially misleading nature. Consequently, without this crucial denominator, it is impossible for external parties to definitively determine whether the per-mile crash rate is improving or worsening.
Lack of Transparency in Incident Reporting
A significant issue highlighted by the data is Tesla's practice of heavily redacting crash reports. The narratives for 99.9% of its 3,763 crash reports have been blacked out, citing confidential business information since 2019. Similarly, the severity of these incidents is often listed as "Unknown," as the underlying telematics data is obscured. The software version involved in crashes is also redacted, making it impossible to differentiate between Autopilot and FSD incidents in the publicly available data. In stark contrast, other major manufacturers like GM, Ford, Honda, and Toyota provide much more detailed, unredacted reports. Tesla has previously argued to regulators that making this information public would cause "financial harm," and the company is currently under a separate NHTSA investigation regarding its crash reporting procedures. Furthermore, reporting delays suggest that the most recent monthly figures are likely understated, with previous "final" May figures having been revised upwards by 139 incidents after new data was added, indicating that the reported 207 crashes are a minimum, not a maximum.
Navigating the Paradox of Improving Technology and Rising Incidents
It is important to acknowledge that Tesla's Full Self-Driving technology continues to evolve and improve. Observers and users have noted the increasing capabilities of the system, recognizing it as one of the most advanced Level 2 systems available in consumer vehicles today. However, this progress introduces a concerning paradox: as the system becomes more adept, drivers may develop a false sense of security, leading to complacency. The marketing of "Full Self-Driving" and promises of imminent autonomy might encourage drivers to pay less attention, treating a Level 2 assist as if it were a fully autonomous robotaxi. This reduced vigilance, combined with improved technology, paradoxically contributes to a record number of incidents, as evidenced by the data. If FSD were truly ready for unsupervised operation, Tesla would likely be aggressively expanding its robotaxi services and deploying unsupervised FSD to consumer vehicles, neither of which is currently happening on a large scale. The company's actions indicate that the technology, while advanced, is not yet fully autonomous, underscoring the critical need to address the discrepancy between marketing and operational reality before further incidents occur.