AI models increase cybersecurity risks for connected vehicles

Peter Ludwig of Applied Intuition warned that the rapid advancement of AI models has made it easier than ever to hack software systems, creating a daily news cycle of large-scale breaches. He cited the recent Binance hack, estimated at $6 billion, as an example of a sophisticated attack that would not have been possible without modern AI models. Ludwig highlighted that many software systems are too static and difficult to update, making them vulnerable if known vulnerabilities exist. He discussed the geopolitical dimension, noting that the US Department of Commerce has banned software and hardware from China and Russia in connected cars due to fears of remote control by attackers. Ludwig confirmed it is technically possible for a political adversary to stop all cars on the street if they run a compromised operating system. He advised organizations to scrutinize their software supply chains for potential malicious incentives or intentional backdoors, particularly given the increased scrutiny on data sovereignty and the US Cloud Act's ability to demand data from US companies.