What was presented
South Korean company CSL (씨에스리, CEO Lee Chun-sik) announced on October 29 that it had shown two systems at KPEX 2026, the international security industry fair held at Songdo Convensia, Incheon: the AI agent for the police called 'Everyone's Police AI' and the investigative data analysis platform 'CyberCop'. The news is reported by ETNews.
According to the company, 'Everyone's Police AI' was developed together with the National Police Agency and includes services for citizens and officers: drafting responses to requests, support for operators of the 182 civil call center, and the chatbot of the Police Civil Service 24 portal. The company says it has structured laws, court rulings and operational guidelines into an industry knowledge base and built a dedicated lexicon, applying retrieval-augmented generation (RAG) to indicate, alongside each response, the rules and sources applied.
Testing status and precautions
CSL maintains that the system has been in internal testing since September on the police network reserved for field officers and that at the fair it received positive assessments of its practical usability. However, these are company statements reported by the source: no quantitative results, user numbers or outcomes of an independent evaluation are provided.
The technically relevant point is the RAG approach: an agent that cites rules and sources can be verified by its users, but quality remains tied to the updating of the regulatory base and to the accuracy of source selection. The material does not clarify who validates the responses or what the final role of human decision-making is in the described flow.
CyberCop and the expansion plan
The second system, 'CyberCop', is described as a platform that combines heterogeneous data—information from open sources (OSINT), account transactions and movement data—to derive investigative clues about intertwined online and offline crimes, such as cyber fraud and phishing. It is under development as part of a national research and development project coordinated by the Science Policing Promotion Center at Hallym University; CSL is responsible for the data platform that integrates different sources and for the web analysis component.
The company, which presents itself as specializing in AI and data for the public sector, declares its intention to extend its business to AI and data for security and justice: to use the knowledge base and operational criteria tested with 'Everyone's Police AI' to bring AI agents into other criminal justice institutions; to develop crime analysis that cross-references investigative and financial data; to transfer data standards and service models developed through contracts with the South Korean police to Asian security agencies, starting with the Philippines.
In support of the strategy, the source reports a statement by Jeon Hye-kyung, CSL vice president: in sectors governed by rules and procedures such as security and justice, the success of AI would depend not on the model but on understanding the institution's work and data and on the ability to redesign that work with AI; the stated goal is an AI in which the machine searches for sources and the person decides. This is a company statement, not a verified fact.
