{"id":1882,"date":"2025-08-29T12:16:58","date_gmt":"2025-08-29T10:16:58","guid":{"rendered":"https:\/\/ai4cyber.eu\/?p=1882"},"modified":"2025-09-09T09:48:04","modified_gmt":"2025-09-09T07:48:04","slug":"ai4cyber-blogpost-security-of-artificial-intelligence-by-ai4cyber","status":"publish","type":"post","link":"https:\/\/ai4cyber.eu\/?p=1882","title":{"rendered":"AI4CYBER Blogpost: Security of Artificial Intelligence by AI4CYBER"},"content":{"rendered":"

Author: Erkuden Rios, TECNALIA<\/strong><\/em><\/p>\n

Artificial Intelligence (AI) systems are increasingly targeted by adversarial threats such as data poisoning and model manipulation. The European Union\u2019s AI Act and ENISA\u2019s threat landscape emphasize the need for AI systems to be safe, transparent, and ethically sound. The AI44CYBER solution for AI trustworthiness, TRUST4AI, supports compliance with these regulations and promotes human oversight.<\/span>\u00a0<\/span><\/p>\n

In this post we summarize how TRUST4AI can support AI risk assessment, with a focus on TRUST4AI.Security, which goal is to prevent adversarial attacks on AI systems.<\/span>\u00a0<\/span><\/p>\n

In the context of the research on how to ensure AI trustworthiness, AI4CYBER developed a security risk assessment methodology for AI systems available at: <\/span>https:\/\/ieeexplore.ieee.org\/document\/10475824<\/span><\/a> , and at: <\/span>https:\/\/zenodo.org\/records\/11209106<\/span><\/a>\u00a0<\/span>\u00a0<\/span><\/p>\n

This is a comprehensive methodology for assessing security risks in AI systems, aiming at ensuring their technical robustness and trustworthiness. As AI technologies become integral to modern digital systems, they also expand the attack surface, necessitating specialized cybersecurity approaches.<\/span>\u00a0<\/span><\/p>\n

The <\/span>AI4CYBER risk assessment methodology for AI systems <\/span><\/b>builds on top of the<\/span> KYKLOS 4.0<\/span>1<\/span> project\u2019s risk assessment methodology for security and privacy of industrial solutions, which has been specified and extended to address specific AI systems related threats, adversarial testing techniques, and mitigations.<\/span>\u00a0<\/span><\/p>\n

The methodology builds on existing frameworks like:<\/span>\u00a0<\/span><\/p>\n