Delivered to more than 40 EU CyberNet Expert Pool members on, the crash course offered practical and exercise-driven approach to realistic OSINT tasks. Explaining from search and aggregation to AI-assisted investigation and validation, the course demonstrated both the power and limitations of AI and highlighted the importance of human-in-the-loop approach. 

Additionally, the course offered an overview of OSINT evolution from manual individual searches, through manual aggregation and cross-referencing, to AI-reasoned and automated searches, and explored the change of human role in AI OSINT: from performing searches individually to framing questions, directing AI-assisted research, assessing sources and validating findings. 

Participants examined how AI can force-multiply OSINT work by exploring more sources, testing hypotheses, identify connection and process information at scale. Course also introduced the emerging OSINT technology stack: from large language model (LLM) chatbots and AI research assistants to agentic systems, capable of carrying out multi-step investigations.  

Furthermore, the course also addressed the risks of AI-assisted OSINT such as hallucinations, fabricated or misleading sources, lack of context or privacy concerns. It was highlighted that the central principle is that AI output is a lead, not evidence.  

As the cyber landscape continues to evolve, continuous learning and initiatives like the Crash Course help EU CyberNet Expert Pool members to stay up to date with emerging threats, technological developments or regulatory changes while exchanging knowledge and exploring new approaches to cyber capacity building.  

Crash Course was led by Ago Ambur, a seasoned Estonian cyber expert, currently working as a Chief Operating Officer with Glazer Technologies and who previously led the Cybercrime Bureau with the Estonian Police and Boder Guard Board.