From 1 to 4 September 2026, the OSCE Programme Office in Astana, in close co-operation with the Law Enforcement Academy under the Prosecutor General's Office of the Republic of Kazakhstan and the OSCE Secretariat's Transnational Threats Department, organized a four-day training course on the Responsible and Ethical Use of Artificial Intelligence in Law Enforcement. The course was conducted within the framework of OSCE activities aimed at strengthening the capacities of law enforcement and specialized agencies to combat transnational crime.
As rapid digital transformation reshapes both the threat landscape and the response toolbox for preventing violent extremism and radicalization that lead to terrorism, with malicious actors increasingly exploiting AI to generate deepfakes, amplify extremist narratives, and obscure identities for recruitment and operational planning, equipping law enforcement to use AI safely and critically has become a practical necessity.
"For law enforcement agencies, new technologies open up significant opportunities - from analysing large volumes of information and documents to improving the efficiency of analytical and other day-to-day work. At the same time, the use of artificial intelligence in such a sensitive area requires particular responsibility. New technologies must not only increase the efficiency of law enforcement activity, but also be applied with proper safeguards for data protection, human rights, and human oversight", said Ambassador Alexey Rogov, Head of OSCE Programme Office in Astana.
Over four days, 24 participants gained an understanding of key AI concepts, including machine learning, large language models, and the difference between open-weight and commercial closed models. They also examined safeguards for the responsible use of AI in law-enforcement settings. Sessions introduced the INTERPOL-UNICRI Toolkit for Responsible AI Innovation in Law Enforcement and discussed how principles such as human oversight, accountability, transparency, and documentation can guide the responsible use of AI.
Participants also practiced the safe use of AI tools for everyday tasks, while discussing their practical limits, the importance of verifying outputs, and the risks of using sensitive or confidential data with public cloud services. Sessions also introduced key technical concepts behind AI systems and explored different deployment options, including how organizations can build internal tools grounded in their own documents.
The course concluded with a group exercise where participants worked through potential AI use cases relevant to their daily work and discussed a roadmap for the responsible adoption of artificial intelligence by law enforcement and specialized agencies in Kazakhstan.
The training was supported under the OSCE Secretariat's Transnational Threats Department'sProject CLEAR (Critical Thinking, Information and Media Literacy, Ethical AI, and Resilience for Preventing and Countering Violent Extremism and Radicalization that Lead to Terrorism), funded by Germany, Finland, Norway, and the United States.
Project CLEAR supports participating States in strengthening comprehensive and human-rights-compliant approaches to preventing and countering violent extremism and radicalization that lead to terrorism that manifest in, and stem from, online environments and the misuse of the internet, emerging technologies, and AI for terrorist purposes across the OSCE area.
