Across the globe, AI-driven audio and video deepfake campaigns in recent years have spread dangerous disinformation on pressing issues such as elections, public health, and climate. And the threat is only growing, as tools capable of producing convincing AI simulations are increasingly accessible and inexpensive, and even simple “cheapfakes” continue to deceive significant numbers of people. In this practical workshop, digital verification journalists lay out the latest tools to track the origins of AI-manipulated content, address the limitations of these tools and techniques to obtain more accurate results.