Cybercrime and the misuse of technology are escalating threats to global security, privacy, and societal stability. While secure system design is essential, persistent flaws—such as the Log4j vulnerability or large-scale disinformation campaigns—reveal the limitations of current approaches. The evolving nature of these threats calls for a paradigm shift. Traditional cybersecurity models fall short when confronted with novel, unforeseen attacks under open-world conditions. SHIFT proposes a transformative research program to address this challenge by exploring how systems can anticipate and adapt to the unknown. The project will develop pioneering methods rooted in uncertainty quantification and adaptive learning, offering a framework that navigates the trade-offs between risk management and the cost of querying oracles for information about unknown threats. This approach departs from closed-world assumptions, opening a novel scientific front that bridges machine learning, security engineering, and forensic science. SHIFT aligns with the spirit of the BBVA Foundation’s Leonardo Grants by venturing into unexplored territory at the intersection of theoretical modeling and real-world experimentation. It introduces new scientific principles to enhance the detection and attribution of cyberattacks, pushing the boundaries of secure-by-design systems and contributing to the theoretical foundations of artificial intelligence and criminology. By fostering dialogue between technology, law, and behavioral sciences, this project aims to provide an interdisciplinary toolkit to better understand, prevent, and respond to cyber misconduct—setting the stage for broader societal resilience in the digital age.

Proyecto funded by Fundación BBVA in the 2025 Leonardo Grants for Scientific Research and Cultural Creation call, awarded to Guillermo Suárez-Tangil Rotaeche. (SHIFT, reference LEO25-1-19720).