Summary of Key Points
This news article focuses on Tesla's Full Self-Driving (FSD) feature, highlighting three main issues: First, a dangerous incident in Canada where a Tesla driver fell asleep while the FSD system failed to detect it, exposing vulnerabilities in the FSD monitoring mechanism. Second, Tesla is addressing these vulnerabilities by implementing facial recognition technology to prevent unauthorized use and enhance safety. Third, two fatal accidents involving FSD in the United States have drawn regulatory attention, highlighting the challenges of FSD in extreme scenarios and human-machine interaction. Overall, although FSD represents technological progress, issues with safety monitoring and practical application still need to be addressed, and the pressure for legal responsibility and regulation is increasing.
1. The Canadian "Sleep-Driving" Incident: Why Did FSD Monitoring Fail?
On a winding road on Highway 1 in Canada, a Tesla driver fell asleep while wearing dark sunglasses. The FSD system did not stop the vehicle until a passerby called the police. Why didn't the monitoring mechanism work?
- Visual Monitoring was Hindered: Tesla's FSD relies on cameras inside the car to monitor the driver's face and eyes to determine if they are awake. However, the driver's sunglasses blocked their eyes, preventing the cameras from detecting the closed-eye state, thus not triggering an alert.
- Physical Monitoring was Interfered with: The system also uses the "torque" on the steering wheel (i.e., whether the driver is actively holding it) to assess attention. If a steering wheel cover is installed, it may prevent the system from detecting the driver's grip, leading it to assume the driver is still in control.
These two vulnerabilities combined caused the system to mistakenly believe everything was normal, and the vehicle continued to drive on the dangerous road.
2. Tesla's "Patch": Facial Recognition to Control FSD Usage
To fix these issues, Tesla has incorporated facial recognition code in its latest app update. This feature serves two main purposes:
- Passenger Verification for Robotaxi: In the future, Tesla's autonomous taxis (Robotaxi) can use facial recognition to confirm passenger identities and prevent unauthorized passengers.
- Authorization Control for FSD: If you have subscribed to FSD, only you (or an authorized person) can use it while driving. This prevents unauthorized use, such as lending the car to someone without experience.
This represents an upgrade in Tesla's monitoring system: from using in-car footage in 2021 (to prevent passenger sabotage) to integrated monitoring in 2023, and now adding facial recognition for more precise control over FSD usage.
3. The Two Fatal Accidents in the United States: What Went Wrong with FSD?
While the Canadian incident did not result in casualties, two fatal accidents involving FSD occurred in the U.S.:
- Arizona Accident: In November 2023, an FSD-enabled Tesla hit a person standing by the side of the road on the highway. The cause was excessive sunlight that blinded the cameras, preventing them from detecting the person and not alerting the driver to take over control, resulting in death. As a result, U.S. regulatory authorities (NHTSA) have intensified their inspection of FSD systems, focusing on the shortcomings of "pure visual" approaches (relying solely on cameras without radar) in bright light conditions.
- Texas Accident: In June 2026, an FSD-enabled Tesla crashed into a house, killing a person. The driver claimed it was due to a FSD malfunction that caused the vehicle to accelerate, but regulatory authorities are also investigating whether there were issues with the system's interaction design, such as the possibility of the driver accidentally stepping on the accelerator in a panic. The driver has been charged with manslaughter.
These accidents highlight significant challenges with FSD in extreme conditions and human-machine coordination.
4. Who Is Responsible When FSD Causes an Accident?
FSD is classified as Level 2 assisted driving, meaning the driver is legally responsible for the accident. However, Tesla cannot escape regulation:
- Canadian Law: Although Level 2 driving is legal, the driver must be awake and hold the steering wheel. In this incident, the driver violated the rules by sleeping, but since Tesla's monitoring system failed to detect it, the company also bears some responsibility.
- U.S. Accidents: The Tesla involved in the Arizona accident settled out of court (the amount is not disclosed); the driver in the Texas accident has been charged, and regulatory authorities are still examining the vehicle's interaction design.
In summary, if a driver uses FSD while asleep or accidentally steps on the accelerator, they are responsible. However, if there are system flaws (such as failing to detect a person in bright light), Tesla will also face regulatory scrutiny, which may lead to recalls or system updates.
Conclusion
Although Tesla's FSD sounds advanced, it still has many safety vulnerabilities. The incidents in Canada and the U.S. serve as reminders that autonomous driving does not mean complete hands-off operation; drivers must remain alert, and car companies need to continuously fix these flaws. Regulatory authorities must also keep up to ensure that technology is safely implemented. When using FSD, users should not assume it will completely free them from responsibility—after all, the current technology has not yet reached that level of reliability.