Artificial intelligence in public administration is often associated with visions of the future or concerns about automated decision-making. In reality, however, its use is far more practical. The presentation will introduce specific projects from public administration and regulated organisations where AI is already helping civil servants, analysts and citizens work more efficiently with large volumes of information. Using real-world examples from ministries, customs administration and other organisations, it will show how AI can support work with extensive case files, document and recording analysis, contract performance monitoring and communication with citizens. The presentation will also highlight that the greatest value of AI in public administration does not come from automating decisions, but from supporting people in making them. It will demonstrate how AI can be introduced safely and transparently in environments with strict requirements for legislation, auditability and data protection.