A brand new EU-funded undertaking, SHASAI, will deal with cybersecurity dangers in AI programs, from design to real-world operation.
Funded beneath the Horizon Europe programme, SHASAI goals to strengthen the safety, resilience, and trustworthiness of AI-based programs as cybersecurity threats change into extra frequent.
It would deal with cybersecurity dangers from the preliminary design and improvement levels by way of to deployment and real-world operation.
Leticia Montalvillo Mendizabal, Cybersecurity Researcher at IKERLAN and SHASAI Undertaking Coordinator, defined: “With SHASAI, we purpose to maneuver past fragmented safety options and deal with AI cybersecurity as a lifecycle problem.
“By combining safe {hardware} and software program, risk-driven engineering and real-world validation, the undertaking will assist organisations deploy AI programs that aren’t solely progressive, but additionally resilient, reliable, and compliant with European rules.”
Utilizing real-world situations to guard AI programs from cybersecurity threats
AI programs have gotten more and more advanced, interconnected, and data-driven, making them targets for cyber-attacks and susceptible to failures that may have real-world penalties.
The undertaking will reveal and validate its strategies and instruments in three real-world situations to assist sort out cybersecurity threats:
- AI-enabled chopping machines within the agrifood sector
- Eye-tracking programs utilized in assistive healthcare applied sciences
- A tele-operated last-mile supply automobile within the mobility sector
These numerous use instances will allow the undertaking’s researchers to check their strategy in several fields whereas guaranteeing that outcomes will be transferred to different AI purposes.
The anticipated end result is a strong, adaptive, and reliable safety structure that ensures AI programs stay resilient, traceable, and compliant with evolving cybersecurity requirements in high-risk environments.
Supporting Europe’s ambitions to advertise reliable AI
By translating high-level cybersecurity and AI security ideas into concrete technical practices, SHASAI additionally helps Europe’s broader efforts to advertise reliable AI.
The undertaking aligns with key EU frameworks and initiatives, together with the EU AI Act, the Cyber Resilience Act (CRA), the NIS2 Directive, and the EU Cybersecurity Technique.
The consortium combines experience from analysis organisations, universities, trade, and know-how suppliers. It began on 1 November 2025 and can run till the top of April 2029.
