
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Researchers at Ben-Gurion University demonstrated that AI-powered autopilot and ADAS systems in vehicles like Tesla and Mazda can be tricked by projected images, causing dangerous maneuvers such as sudden braking or swerving. This vulnerability could be exploited remotely, posing a significant safety risk, though no real-world incidents have occurred yet.[AI generated]
Why's our monitor labelling this an incident or hazard?
The event involves AI systems explicitly (Tesla autopilot and Mobileye ADAS) whose malfunction or exploitation via 2D projections can cause dangerous driving behaviors. While no actual incident of harm is reported, the demonstrated ability to trick these systems plausibly leads to safety risks and potential injury or death. The researchers notified manufacturers but no fix has been implemented yet, underscoring the ongoing hazard. This fits the definition of an AI Hazard as it plausibly could lead to an AI Incident involving injury or harm to persons.[AI generated]