Busan Police Launch AI Training to Automate Reports and Predict Crime

The Busan Metropolitan Police Agency held its first formal artificial intelligence training session on May 6, 2026, marking a shift toward what officials call "smart policing." The session focused on practical applications of generative AI and data-driven technologies in actual police work, from automating routine reports to analyzing crime patterns.

According to Asia Business Daily, the training theme was "The Era of Smart Policing Opened by AI Agents and Vibe Coding." Professor Noh Cheolwoo, an advisory professor at the Busan Science and Technology Advisory Group, delivered the lecture. With over 30 years of research experience in AI and digital transformation, Noh presented practical cases for police work applications.

The training covered several concrete topics. Officers learned about automating report writing, improving case law and statute search efficiency, crime prediction systems, and AI ethics standards. The session introduced that AI can reduce the time required for report writing by up to 50%. The intention is to minimize repetitive administrative work so that officers can focus more on field response (which, let's be honest, is where they actually want to be).

International examples were shared during the session. The training explained trends in data-driven preventive policing by referencing the crime prediction system in the United States and the recidivism risk analysis model in the United Kingdom. The use of intelligent CCTV and data analytics to detect early signs of crime was highlighted as a key topic.

Professor Noh stated, "AI is not a technology that replaces the police, but a tool to support their work," emphasizing, "The final judgment and responsibility remain with the police officer." This distinction matters. When an officer sits at a desk typing incident reports at 2 a.m., the physical reality involves tired fingers, dim screens, and the mental fatigue that comes from repetitive data entry. AI tools aim to reduce that friction, but the officer still signs off on every decision.

The training also stressed the importance of personal data protection and security issues. Practical guidelines were provided, including the principle of not inputting suspect information or investigative secrets into AI systems and the necessity of verifying AI-generated results. This is critical because public sector AI deployment carries different risks than commercial applications.

A representative of the Busan Metropolitan Police Agency said, "Our goal is to enhance administrative efficiency and strengthen citizen safety through the adoption of AI," adding, "We plan to continue expanding training programs to further develop digital-based policing capabilities." The agency intends to gradually broaden the scope of AI training and technology utilization by collaborating with local experts.

This police training operates within a broader municipal framework. The Busan Metropolitan City announced its "2026 Busan Metropolitan City Artificial Intelligence Administration Implementation Plan" to systematically introduce AI across all areas of administration. The plan focuses on enhancing administrative work efficiency through AI technology and expanding services citizens can experience in daily life.

Under the vision of "A Global AI-Powered Intelligent Administration City Growing Together with Citizens," the city set two key priority tasks: AI-based intelligent administrative innovation and expansion of empathetic AI that transforms daily life. The plan includes 38 detailed tasks to be implemented in phases. In the safety sector, the city will expand AI use in areas directly related to citizen safety, including AI safety monitoring for fire prevention in traditional markets and intelligent traffic management that predicts congestion and accident risks.

The implementation plan is a comprehensive annual plan formulated pursuant to Article 5 of the "Ordinance on the Creation of Artificial Intelligence Administration of Busan Metropolitan City." It focuses on enhancing the efficiency of administrative work through the use of AI technology and expanding administrative services that citizens can experience in their daily lives. The city will also foster practice-oriented AI talent by strengthening AI competency training for public officials and operating AI study groups.

Whether this training translates to measurable improvements in police work remains to be seen. The 50% reduction in report writing time is a claim that needs real-world validation across different precincts and case types. Some officers may adapt quickly to AI tools while others resist the change. The technology itself is only as useful as the people wielding it.

Security concerns also persist. The guideline against inputting suspect information into AI systems creates a practical constraint—officers must constantly evaluate what data can be processed. This adds cognitive load to an already complex workflow. The friction between efficiency gains and security protocols will likely surface in daily operations.

Busan's approach positions it ahead of some regional peers in AI adoption for public safety. However, the real test comes when officers face actual incidents and must decide whether AI recommendations align with their professional judgment. The training provides tools, but it cannot replace experience or accountability. Whether citizens notice any tangible improvement in response times or safety outcomes is the question that matters most.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn