Tesla delays driverless car plans to mid-2027, limiting use to small robotaxi zones

Tesla has officially pushed its target for unsupervised Full Self-Driving in customer-owned vehicles to the second quarter of 2027. This announcement, made during the company’s third-quarter 2026 earnings call, marks the latest adjustment to a promise first made nearly a decade ago. However, the new timeline comes with a significant restriction: the feature will not be a general software update that allows owners to drive anywhere, but rather a mechanism for adding private vehicles to Tesla’s existing robotaxi network.

The clarification came from Ashok Elluswamy, Tesla’s AI chief, who responded to a high-priority question from retail shareholders. The query, supported by approximately 1.1 million shares, asked when unsupervised FSD would become available in personal cars. Elluswamy’s response indicated that the initial rollout will function as an integration into Tesla’s managed fleet. This means a car will only drive itself without human supervision within the specific geofenced areas where Tesla’s robotaxi service is already operational.

These operational zones are currently limited to six metropolitan areas in the United States, and even within those cities, the service areas are relatively small. For instance, the Miami zone primarily covers West Miami and a narrow strip extending toward Doral and Sweetwater. Similar limited geofences were established in Orlando and Tampa in July. The service generally adheres to low-speed routes, avoids highways, and steers clear of complex urban environments. In the San Francisco Bay Area, the system is not fully unsupervised due to California regulations that require a safety driver unless specific data is released to the Department of Motor Vehicles, a step Tesla has delayed.

The scale of the current robotaxi fleet remains modest. Tesla reported that the service accumulated nearly two million unsupervised miles by the end of the third quarter, a month after the Cybercab model was added to the Austin fleet. Third-party tracking data from earlier this year suggested that only dozens of fully driverless Teslas were on the road, a figure far below the hundreds of thousands previously cited by CEO Elon Musk. Competitor Waymo, which entered Orlando and Tampa before Tesla, has already logged over three million miles.

This current approach contrasts sharply with the original marketing pitch. For years, Tesla sold FSD as a technology that would eventually allow any car to drive itself anywhere via a software update. In July 2025, Musk stated that unsupervised FSD would be available for personal use in certain geographies by the end of that year. A year prior, he expressed surprise if the technology was not ready the following year. Complicating the issue further, owners of hardware version 3 (HW3) have been informed that their vehicles cannot run unsupervised FSD at all, with no clear upgrade path provided.

The shift in strategy suggests a more conservative rollout than previously advertised. Instead of a universal capability, the product is now defined by its integration into Tesla’s fleet, operating within Tesla’s chosen zones under Tesla’s terms. This excludes long-distance highway trips and cross-town commutes unless the owner resides and works within one of the limited service areas. The change effectively moves the goalpost from a car that drives the owner anywhere to a car that can function as a robotaxi within a controlled environment.

During the earnings call, Elluswamy identified safety as the primary bottleneck for expansion. He suggested that the company is limiting operations to low-speed roads and simpler environments to manage risk, with the hope that these safety concerns will be resolved by the following year. However, he acknowledged uncertainty regarding the timeline. The move to Q2 2027 represents the latest iteration of a date that has been repeatedly adjusted since 2019, leaving many long-time FSD buyers with a narrower product than the one originally promised.