Azerbaijan's cyber security agency drafts AI-use rules for government bodies

Azerbaijan's State Service for Special Communications and Information Security (Xüsusi Rabitə və İnformasiya Təhlükəsizliyi üzrə Dövlət Xidməti) has begun drafting requirements for how government agencies use and implement artificial intelligence, department head Nadir Novruzov said at a panel discussion during ADEX 2026, the 6th Azerbaijan International Defence Exhibition, Report.az reported on 2 October 2026. News.az confirmed the same remarks the same day.

Rules on paper, AI already in use

• Novruzov said the agency has been granted authority to identify potential risks associated with AI, according to Report.az.
• An existing annual summit of information security and IT heads from government agencies will hold its ninth edition on 20 October, with AI security as its main focus, Novruzov said; official invitations to agencies are due to follow, per Report.az and News.az.
• In a separate Report.az article from the same panel, Novruzov said the agency is already using AI to check internet-accessible government information systems, because "there are hundreds or even thousands" of them and manual checks cannot cover them all.
• "We compare the results of audits and penetration tests conducted by humans with AI-generated results and improve the missing parts so that, in the future, penetration tests can be carried out using artificial intelligence," Novruzov said, according to Report.az.
• He grouped the agency's AI-security concerns into two areas: the security of AI tools deployed by agencies themselves (data leaks, prompt injection, data poisoning) and the risk of staff leaking confidential information through public tools such as ChatGPT and Claude, Report.az said.

Why it matters

A state cybersecurity body setting rules for AI use is a governance step, not a ban: Report.az's account describes risk-mapping and a forthcoming summit, with no published text of the rules themselves or a date for one. The parallel disclosure that the same agency already runs AI-assisted penetration testing on government systems shows Azerbaijani institutions adopting AI operationally before the rulebook is finished, a sequencing regulators in the EU and elsewhere have also struggled to avoid.

What it means for Azerbaijani companies

Azerbaijan has an Artificial Intelligence Strategy for 2025–2028, which Report.az describes as the framework state AI projects are aligned with. Vendors selling AI-enabled software, cybersecurity tools or cloud services to Azerbaijani state bodies should expect the 20 October summit and any resulting requirements to shape procurement conditions, particularly around data handling and the use of public chatbots by government staff; Report.az did not report a timeline for when formal rules would take effect.