Tesla hits Musk’s threshold for ‘safe unsupervised’ driving

2026-05-04 · Source: The Verge · Field: Technology & Digital — Robotics & Autonomous Systems, Artificial Intelligence & Machine Learning · Depth: Fundamental Awareness, quick

Summary

Tesla's Full Self-Driving (Supervised) system has accumulated over 10 billion miles, a milestone Elon Musk previously stated was needed for "safe unsupervised" driving. Despite reaching this threshold, FSD remains a Level 2 system requiring active human supervision, and Tesla has not enabled unsupervised driving for its customer fleet. The company faces ongoing questions regarding liability for crashes involving its partially autonomous features, having settled hundreds of cases and been found partly liable in a 2019 Autopilot crash, ordered to pay $243 million. While Tesla claims FSD-equipped vehicles drive 5.5 million miles before a major collision compared to 660,000 for average US drivers, experts question its safety reporting methodology. Tesla is, however, expanding its unsupervised robotaxi fleet in cities like Austin, Dallas, and Houston.

Key takeaway

For AI Product Managers evaluating autonomous driving features, understand that achieving mileage milestones does not automatically translate to higher autonomy levels or reduced liability. Your product's legal and regulatory framework, particularly concerning accident responsibility, is as critical as its technical performance. Ensure your roadmap accounts for legal clearances and liability shifts before promising unsupervised capabilities to customers, as Tesla's experience shows these are significant hurdles.

Key insights

Tesla's FSD (Supervised) has logged 10 billion miles, yet unsupervised driving for customers remains elusive due to liability concerns.

Principles

• Liability shifts with autonomy levels.
• Data volume does not equate to regulatory approval.

In practice

• Monitor Tesla's robotaxi fleet expansion.
• Review FSD terms of service for liability clauses.

Topics

• Tesla Full Self-Driving
• Autonomous Vehicle Liability
• Self-Driving Milestones
• Level 2 Autonomy
• Robotaxi Fleet

Best for: CTO, Executive, AI Product Manager, Tech Journalist, Legal Professional, General Interest

Related on AIssential

• Tesla’s Robotaxis Are Going Nowhere — Tesla's robotaxi claims are unsubstantiated by actual autonomous vehicle testing in California.
• Why it might not make sense for you to own a self-driving car — Consumer ownership of fully self-driving cars faces significant hurdles due to extreme hardware costs and complex maintenance requirements.
• The Road Ahead for Autonomous Vehicle Regulation — Federal and state governments should divide autonomous vehicle regulatory responsibilities to balance safety, innovation, and local needs.
• Why “Physical AI” Is Quietly Overtaking Tesla — And the Market Is Still Asleep — Distributed, infrastructure-aware "Physical AI" architectures are outperforming Tesla's closed, vehicle-centric autonomous driving system.
• Elon Musk admits millions of Tesla owners need upgrades for true ‘Full Self-Driving’ — Tesla's Hardware 3 is insufficient for unsupervised Full Self-Driving, necessitating extensive upgrades.
• Tesla reveals two Robotaxi crashes involving teleoperators — Teleoperator intervention in Tesla Robotaxis has led to low-speed crashes, revealing safety challenges in autonomous vehicle deployment.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Verge.