U.S. Senator Demands NHTSA Verify Tesla FSD Safety Data Accuracy and Questions Its Methodology

U.S. Senators Ed Markey and Richard Blumenthal have sent a letter to the National Highway Traffic Safety Administration (NHTSA), urging the agency to verify whether Tesla’s publicly released data on the safety performance of its Full Self-Driving (Supervised) system is accurate and reliable. Tesla has claimed that vehicles equipped with FSD experience one serious crash per approximately 5.5 million miles driven, compared to human drivers, who average one such crash every 660,000 miles—asserting that FSD is seven times safer than human driving. The senators warned that if these statistics are misleading, they could lead drivers to over-rely on the system, obscure potential risks, and impair regulators’ ability to assess vehicle safety. They requested that NHTSA respond by July 7 to four specific questions: whether it has independently verified Tesla’s safety conclusions; whether it has ever required Tesla to provide raw data, underlying assumptions, definitions of crashes, and complete methodologies; whether it has evaluated the impact of Tesla’s use of a 5-second window—instead of the industry-standard 30 seconds—for counting disengagements or takeovers; and whether it has reviewed potential gaps in Tesla’s reliance on connected onboard devices to automatically collect crash data. Previously, Reuters raised concerns about the rigor of Tesla’s statistical methodology. Earlier this year, California’s Department of Motor Vehicles determined that the name “Autopilot” was misleading, prompting Tesla to discontinue its use. Meanwhile, the Netherlands’ transport authority approved FSD for use based on independent real-world testing data from the Dutch Vehicle Authority (RDW), stating that the system had accumulated 24 million kilometers (approximately 15 million miles) of driving in the country without a serious crash—making it 3.5 times safer than human drivers. As federal scrutiny intensifies, New Jersey is currently considering two bills (S1677 and H3968) that would impose restrictions on autonomous vehicle testing and operations. Tesla has launched lobbying efforts against the legislation, arguing that the proposed requirements set excessively high barriers that hinder the legal deployment of fully driverless vehicles and could create exclusionary industry barriers. In contrast, Texas has already allowed Tesla to complete Level 4 autonomous driving compliance self-certification and has permitted the company to operate driverless robotaxi services without safety drivers in Dallas and Houston. NHTSA confirmed receipt of the senators’ letter and has initiated a review. Individual states retain independent regulatory authority over the deployment of Tesla’s autonomous vehicles.