Are autonomous vehicles really smart? Check how a simple trick can fool them
2024-05-06
https://auto.hindustantimes.com
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Arizona content creator Jason Carr demonstrated that Waymo's autonomous vehicles can be tricked into stopping by wearing a t-shirt with a stop sign image. The AI system misinterpreted the shirt as a real traffic signal, revealing a vulnerability that could lead to traffic disruptions or safety risks if exploited.[AI generated]
Why's our monitor labelling this an incident or hazard?
The event involves an AI system (autonomous vehicle) whose use (perception and decision-making) was directly affected by a simple trick, causing it to stop unnecessarily. While no actual harm (injury, property damage, or rights violation) is reported, the incident demonstrates a malfunction or failure to correctly interpret inputs, which could plausibly lead to safety risks or operational disruption if exploited or occurring in critical situations. Therefore, it qualifies as an AI Hazard because it plausibly could lead to harm, but no harm has yet occurred as per the article.[AI generated]