Tesla Publishes New FSD Safety Data

Tesla has published a new Full Self-Driving (Supervised) Vehicle Safety Report, expanding the company’s public data on how its driver-assistance system performs compared with manually driven Tesla vehicles. Tesla says the report uses collision data collected through vehicle telemetry and compares FSD Supervised with manual driving, including vehicles with and without active safety features.

The newly released European data covers more than 100 million kilometres of driving by over 70,000 Tesla customers, according to reports on the dataset. Tesla claims that vehicles using FSD Supervised were approximately 4.1 times less likely to be involved in a collision than Tesla vehicles driven manually. The publication comes as European regulators continue considering the wider deployment of Tesla’s advanced driver-assistance technology.

Tesla argues that publishing real-world driving data provides evidence that its AI-powered driving system can improve road safety. However, the company still classifies FSD as a supervised driver-assistance system, meaning drivers remain responsible for monitoring the vehicle and taking control when necessary. It is not equivalent to fully autonomous driving.

The data release has also revived questions about how Tesla measures and presents safety performance. Earlier Reuters reporting found that Tesla had presented self-published safety statistics to European regulators, while independent traffic-safety researchers raised concerns about whether some comparisons could create a misleading impression of the system’s performance.

The debate highlights a major challenge for the autonomous vehicle industry: proving that AI-driven systems are genuinely safer requires large datasets, transparent methodology and independent evaluation. Tesla’s latest report adds more information to the public discussion, but regulators and safety experts are likely to continue demanding independently verifiable evidence as FSD expands into more markets.