The updated safety page says the supervised system has exceeded Elon Musk's January threshold for “safe unsupervised” driving, but customer cars remain Level 2 vehicles requiring constant human attention.
Musk said last month unsupervised driving in customer cars could arrive in the fourth quarter when legal, while liability for crashes remains unresolved because Tesla's terms place responsibility on owners.
Tesla is meanwhile expanding robotaxis, with five unsupervised vehicles in Dallas, six in Houston, and in Austin 22 unsupervised and 29 supervised vehicles, as critics continue to challenge its safety claims.
Tesla's robotaxis run unsupervised, so why are customers still legally liable for the same FSD software in their own cars?
After 10 billion miles, why is Tesla's FSD crash rate in Austin still four times higher than human drivers?
With 4 million cars needing hardware upgrades for FSD, what does Tesla's promise of full autonomy mean for its earliest buyers?