New AI Technique Sounding Out Audio Deepfakes

Summary

Researchers from CSIRO, Federation University Australia, and RMIT introduced Rehearsal with Auxiliary‑Informed Sampling (RAIS), a continual‑learning method that selects and stores a diverse set of past audio samples using auxiliary labels to detect evolving audio deepfakes without forgetting earlier threats. The technique achieved a low 1.95% error rate across multiple attack scenarios, demonstrating superior adaptability compared to traditional detectors. Dr. Kristen Moore highlighted RAIS’s ability to retain hidden audio traits, while PhD graduate Falih Gozi Febrinanto emphasized its practical impact for safeguarding voice‑based authentication and combating disinformation.