Author: Valentina Del Rio / PLURIBUS
During the final event of #LAZARUS_EU, we had the opportunity to share the work behind KINAITICS, a Horizon Europe project that explores how artificial intelligence is increasingly becoming a key player not just in defending cyber-physical systems, but also in threatening them.
Valentina Del Rio, Research & Innovation EU Project Manager at Pluribus One, took the stage to explain how AI, while often seen as a security ally, can also be exploited to manipulate robots, drones, and industrial control systems. The KINAITICS project is all about understanding these emerging risks and developing practical, forward-looking defenses.
To do so, it challenges the perception systems that control intelligent machines through simulated attacks, pushing them to their limits to uncover where and how they might fail. From there, the project works on building a robust defense framework that can detect anomalies in real time, validate behaviors based on physical models, and initiate automated recovery when things go wrong.
What makes this initiative even more valuable is its openness: researchers will be able to test new protection mechanisms on real industrial hardware through dedicated testbeds, helping bridge the gap between theory and real-world resilience.
A big thank you to all the LAZARUS_EU partners who helped make this session a success. Conversations like this are essential as we continue to shape a safer future in the age of AI-driven systems.