Tech News: Autonomous vehicles may be vulnerable to attack.
Autonomous vehicles promise to make driving safer for all by eliminating the most common cause of road accidents, human drivers. However, such vehicles can pose a very different type of risk to drivers, passengers and even pedestrians.
To drive on their own, autonomous vehicles use artificial intelligence systems that employ machine learning techniques to collect, analyze and transfer data to make road decisions that would normally be made by human drivers. However, like all IT systems, these systems are vulnerable to attacks that could compromise the operation of self-driving vehicles.
The European Cybersecurity Agency (ENISA) and the Joint Research Center (JRC) have published a new report, entitled “Cybersecurity Challenges in the Uptake of Artificial Intelligence in Autonomous Driving”, which highlights the risks associated with the use of AI in autonomous vehicles and makes recommendations to mitigate these risks.
In addition to unintended threats such as sudden failures, these systems are also vulnerable to intentional attacks with the aim of interfering with their artificial intelligence systems. For example, adding paint to a road or stickers to a stop sign could prevent the AI system of an autonomous vehicle from proper operation which could lead to items being classified incorrectly in a way which could result in dangerous vehicle behavior.
In order to improve AI safety in autonomous vehicles, the report recommends that regular safety assessments of AI components be conducted throughout their lifecycle. The report also suggests that risk assessments are undertaken to identify potential AI risks.