A Reuters investigation reveals deep-seated distrust among former Tesla employees regarding the 'Full Self Driving' system (FSD). The news agency interviewed a former autonomous driving engineer and nine former data labelers – with sobering results.
Core Criticism: Data Quality and Safety
Respondents expressed significant concerns about FSD reliability. A central point of criticism was the poor quality of training data. Data labelers reported unclear instructions and inconsistent labeling, which could lead to erroneous learning outcomes. The former engineer also doubted the company's safety philosophy, which relies too heavily on AI promises and too little on proven safety standards.
Comparison of Hardware Generations
Europe vs. USA: Regulatory Hurdles
While Tesla is rolling out FSD as a 'Supervised' system in the USA, EU approval remains pending. The European Commission requires stricter evidence in accordance with UNECE regulations. The criticism from former employees could further unsettle already skeptical European authorities. Read more here: Tesla FSD in Europe: Next Approval in Another Country Imminent.
Impact on Robotaxi Plans
Tesla has announced plans to launch a robotaxi network in the USA starting in 2026. The latest revelations raise the question of whether the system is truly ready for unmanned operation. A detailed look at the challenges can be found in the article Tesla Cybercab: Extremely Efficient – But Questions and Hurdles Remain.
The discrepancy between Tesla's public optimism and internal doubts shows: the path to fully autonomous driving remains rocky – and the problem areas lie not only in regulation but also within its own workforce.