US-led team introduces framework to make networks of robots, self-driving cars safer

Harvard researchers developed a "cy-trust" framework that helps networks of self-driving cars and robots evaluate which information to trust before making critical decisions. The system assigns numerical trust scores to data received from other machines, allowing autonomous systems to ignore potentially harmful or false information from compromised agents. Testing showed that cooperative robots successfully identified and rejected suspicious inputs from simulated attackers while maintaining normal operations, demonstrating improved resilience against cyber threats that could cause real-world accidents.