Description
DEFORM is an EU Horizon research project addressing the growing challenge of deepfakes, AI-generated or manipulated audio, video, and images increasingly used in fraud, misinformation, and crime. These developments are pushing the limits of current forensic methods, creating a clear need for more reliable detection tools.
The project aims to develop a flexible AI framework capable of identifying deepfakes and tracing their origins. A key innovation is the integration of explainable AI (xAI), ensuring that results are not only accurate but also transparent and understandable for investigators, legal professionals, and the public.
By leveraging advanced foundation models and learning from limited data, DEFORM’s solutions will be adaptable to new and evolving manipulation techniques. The technologies will be tested in real forensic environments across Europe.
The consortium includes leading research institutions, law enforcement agencies, and forensic experts from six European countries. Together, they will deliver practical tools for analyzing images, audio, and video, along with user-focused explanations and open resources. Coordinated by Colourlab, DEFORM aims to strengthen the fight against digital crime and build greater trust in AI.
Visit the project website for more information on the project.
