Advances in generative AI led to high-quality audio synthesis models for diverse applications. From Google DeepMind’s WaveNet to Microsoft’s VALL-E, these innovations have permeated various industries. Yet, the surge in realistic audio deepfakes poses risks, exploited by scammers to impersonate voices in scams. In response, deep learning is harnessed for precise detection tools, with ongoing challenges and commercial solutions fortifying against potential misuse. EITHOS pledges to develop robust forensics algorithms to aid law enforcement in combating audio deepfake abuse.