Tesla delays driverless driving for owners to mid 2027 with strict limits

1.1 million shares backed a specific question during Tesla's third quarter 2026 earnings call today: when can customers expect unsupervised Full Self Driving in their personal vehicles? The company answered that this capability is targeted for the second quarter of 2027. This date represents the latest iteration of a promise Elon Musk has made to buyers for nearly a decade, yet the current plan introduces significant geographical and operational constraints that differ from earlier marketing.

Ashok Elluswamy, Tesla's AI chief, clarified the scope of this update on the call. He described the feature not as a software update that allows owners to travel anywhere unassisted, but as a mechanism for owners to add their private vehicles to Tesla's managed Robotaxi fleet. In this model, the car drives itself unsupervised only within the specific zones where Tesla already operates its commercial robotaxi service. This means the vehicle does not gain universal autonomy, but rather joins a limited network controlled by the manufacturer.

The current operational footprint for this service is modest. Tesla's Robotaxi service is active in six U.S. metropolitan areas, but the service areas often cover only small portions of those cities. For example, the Miami zone primarily includes West Miami and a narrow strip extending toward Doral and Sweetwater. The Orlando and Tampa zones, which launched in July, also feature similarly small geofences. Within these boundaries, the vehicles generally stick to low-speed routes and avoid highways and more complex urban environments.

A notable exception exists in the Bay Area, where the service is not yet fully unsupervised. California regulations require a safety driver in every vehicle because Tesla has not released the necessary data to the California DMV to move beyond a level 2 autonomous system classification. This regulatory hurdle highlights the gap between the company's internal testing and its external deployment capabilities in key markets.

The scale of the current unsupervised fleet remains small relative to previous executive statements. Tesla reported reaching nearly 2 million cumulative unsupervised miles by the end of the third quarter. This milestone came one month after the Cybercab joined the Austin fleet. However, third-party tracking data from this spring indicated that only dozens of fully driverless Teslas were actually on the road, a number far below the hundreds of thousands Musk had previously discussed. Competitor Waymo, which entered Orlando and Tampa before Tesla, has already logged over 3 million miles.

This new approach contrasts sharply with the pitch presented to FSD buyers for years. The original promise was a car that would eventually drive itself anywhere via a software update. In July 2025, Musk told investors that unsupervised FSD would be available for personal use in certain geographies by the end of that year. A year prior to that, he stated he would be surprised if Tesla did not have unsupervised self-driving the following year. Furthermore, owners of Hardware 3 vehicles have been informed that their existing computers cannot run unsupervised FSD, with no clear upgrade path provided.

Elluswamy identified safety as the primary bottleneck delaying broader rollout, which is why the system currently limits itself to simpler driving environments. He suggested these safety issues might be resolved by next year, though he admitted the company is still guessing on the exact timeline. For consumers, the shift means the product is now narrower: the car joins Tesla's fleet, operates in Tesla's zones, and runs on Tesla's terms. This excludes highway road trips, cross-town airport runs, or unsupervised commutes unless the driver lives and works strictly inside one of the designated small zones.