A recent report from Gartner indicates that a misconfigured artificial intelligence (AI) system may lead to substantial disruptions in critical infrastructure by the year 2028. This concern highlights the risks of human error during the configuration process of AI systems, particularly in cyber-physical environments, rather than traditional threats like hacking.
Cyber-physical systems, which integrate physical processes with computing and networking, are increasingly dependent on AI technologies for managing functions like electricity generation. Gartner warns that errors in these AI-driven control mechanisms could result in severe damage, operational shutdowns, and disruptions to supply chains. The vice president analyst at Gartner, Wam Voster, stated that the next major infrastructure failure could be due to a well-meaning engineer's mistake, rather than external threats.
Additionally, Darren Guccione, CEO of Keeper Security, supported Gartner's findings, emphasizing that the rapid integration of AI in sectors such as power, healthcare, and transportation is outpacing governance and security measures. He cautioned that misconfiguration remains a significant risk, particularly as automation increases, and pointed out that AI does not mitigate risk but can actually amplify it when proper safeguards are lacking.