Establishing general rules alone is not enough to ensure the safety of artificial intelligence (AI). It is equally important to determine how these rules should be implemented in information systems, how specific security measures should be put into practice, and how their effectiveness should be measured.

A framework built around nine key pillars has been proposed to support the safe and responsible governance of AI in Azerbaijan. It addresses governance and accountability, security, data protection and integrity, transparency and auditing, continuous monitoring, and the governance of autonomous AI systems.

These findings are presented in a research study on the safe and responsible governance of AI conducted by Leyla Dadashova, a researcher at the Institute of Cybersecurity and Defence Technologies of Azerbaijan Technical University (AzTU), under the supervision of Associate Professor Rahid Alekberli.

One of the study’s key contributions is its focus on translating general AI governance principles into concrete security measures. The research also proposes indicators for identifying security incidents and measuring the time required to resolve them. This approach enables the security of AI systems to be evaluated against measurable performance indicators.

Alongside international best practices, the study takes into account Azerbaijan’s existing legal and strategic frameworks. The proposed approach is aligned with the Artificial Intelligence Strategy for 2025–2028, the Information Security and Cybersecurity Strategy for 2023–2027, and the regulatory framework governing the protection of critical information infrastructure.

The study is among the research outputs of AzTU’s Institute of Cybersecurity and Defence Technologies, which conducts research in AI governance and cybersecurity governance. Beyond its research activities, the Institute focuses on integrating the latest technological advances in cybersecurity and defence technologies into academic programmes, developing new teaching and research areas, building and strengthening academic capacity, and fostering scientific, technological and professional expertise in these fields.

The Institute’s core research areas include AI governance, AI ethics and trustworthy AI, the security of AI-enabled and cyber-physical systems, national cybersecurity governance and cyber resilience, and the governance and resilience of critical information infrastructure.

The research conducted by Leyla Dadashova under the supervision of Rahid Alekberli received second place in the Best Paper Award category in the Economic and Financial Security, Financial Intelligence and Management Systems track at the International Conference on Cybersecurity, Digital Forensics and AI Applications, held in Istanbul in June 2026.

The paper has also been published on IEEE Xplore, the digital library of the Institute of Electrical and Electronics Engineers (IEEE).