Former Tesla employees who helped train the company's Full Self-Driving system are openly refusing to ride in vehicles equipped with it, raising serious questions about the gap between Tesla's public safety claims and what insiders actually witnessed during development. According to a Reuters investigation published on May 28, 2026, ten former Tesla employees, including nine data labelers and one former Autopilot engineer, expressed widespread distrust in FSD's safety.

What Are Former Tesla Insiders Saying About FSD Safety?

The concerns from these former employees are striking in their directness. Seven of the nine data labelers explicitly stated they would not ride in a Tesla vehicle with FSD enabled. One former employee said, "We've all seen FSD fail," and directly contradicted Elon Musk's public claims about the system's capabilities. Another former employee was even more blunt, stating they wouldn't ride in a Tesla Robotaxi even if paid to do so.

These workers had spent considerable time reviewing vast amounts of FSD driving footage used to train the software to avoid repeating mistakes. Their frontline perspective on the system's performance carries particular weight because they saw the raw data behind the technology before it reached consumers. According to their accounts, they frequently observed FSD-equipped vehicles speeding during testing, yet engineers and management appeared not to treat such issues as high-priority concerns.

How Does This Compare to Competitors' Liability Approaches?

Tesla's approach to FSD liability stands in sharp contrast to recent moves by competitors. While Tesla continues to stress that drivers must remain attentive and monitor road conditions at all times while using FSD, and has not made any formal, comprehensive compensation guarantee, other automakers are taking different stances. BYD, the Chinese automaker, announced on May 28, 2026, that it will provide 100% compensation for accidents occurring while using its God's Eye Level 2+ Urban NOA (Navigation-Assisted Driving) and Intelligent Parking features, without any cap on payout amounts.

This represents a fundamental philosophical difference. BYD became the first major global automaker to explicitly assume manufacturer liability for accidents occurring during the use of driver-assistance systems. The compensation covers vehicle repair costs, third-party property damage, and legal liabilities arising from personal injury, and is offered free of charge without affecting the customer's insurance premium for the following year. The core significance of BYD's move lies in its willingness to take responsibility for technology failures, whereas Tesla has traditionally emphasized that drivers bear primary responsibility in incidents involving Level 2 and 3 Advanced Driver Assistance Systems (ADAS).

Steps to Understanding the Safety Debate Around Self-Driving Systems

• Level 2 vs. Level 3 Systems: Level 2 systems like Tesla's FSD require drivers to remain engaged and ready to take control at any moment, meaning the driver retains primary responsibility. Level 3 systems allow more autonomous operation but still require driver availability. Understanding this distinction is crucial because it affects who bears liability when accidents occur.
• Manufacturer Liability Models: Traditional automakers have emphasized driver responsibility, arguing that since drivers must monitor the road, they should bear primary liability. BYD's new approach assumes manufacturer responsibility if the system fails while being used correctly, fundamentally shifting the risk model and signaling confidence in the technology.
• Data Labeling and Training: Self-driving systems rely on human workers reviewing thousands of hours of driving footage to label scenarios and train the AI. The quality and thoroughness of this labeling process directly impacts system safety, which is why insider perspectives from former data labelers carry significant weight in assessing real-world performance.

The credibility gap between Tesla's public messaging and what former insiders witnessed during development work represents a significant challenge for the company. While Tesla has positioned FSD as approaching full autonomy, the testimony from employees who directly observed the system's performance suggests the reality may be more complicated. The fact that people who worked on training the system itself express reluctance to trust it as passengers raises questions that extend beyond typical competitive rivalry.

BYD's willingness to back its technology with unlimited manufacturer liability, even if limited to one year and the Chinese market, sends a powerful market signal about technological confidence. Whether this approach will pressure Tesla and other competitors to adopt similar liability models remains to be seen, but it introduces a new dimension to the autonomous driving competition: not just technical capability, but willingness to stand behind that capability with financial responsibility.

The contrast between insider skepticism and public confidence in FSD highlights an ongoing tension in the autonomous vehicle industry. As these systems move closer to widespread deployment, the gap between what developers claim and what those closest to the technology actually believe about its safety will likely become an increasingly important factor in consumer adoption and regulatory approval.