In this episode of Corruption, Crime and Compliance, Michael Volkov breaks down why the EU AI Act has moved from a future planning exercise to an actively enforced regulatory regime, with the European Commission’s AI Office holding full investigative and fining authority since August 2026, having already opened its first formal investigations in June 2026 into hiring tools, credit scoring systems, and student monitoring applications. He walks through the Act’s fragmented compliance timeline, prohibited practices enforceable since February 2025, general-purpose AI obligations running since August 2025, and live chatbot transparency requirements, alongside the significant deadline relief the Digital Omnibus gave specifically to high-risk AI systems, pushed to December 2027. The episode closes with a clear warning: companies that read the Digital Omnibus as a blanket delay of the entire AI Act are making a costly mistake, since the tracks carrying real, current enforcement exposure, including fines up to 7 percent of global turnover, remain fully active today.