Researchers behind a new deepfake detection program say AI-generated faces have become more convincing, making it harder to distinguish them from real people. Both outlets report that the expert characterizes the current situation as a “cat and mouse game,” where improvements in deepfake quality and advances in detection evolve in parallel. According to the reports, the researchers’ work indicates that the increased realism of AI faces can lead observers to doubt real footage as well, shifting skepticism toward genuine content. The coverage focuses on the detection challenge created by rapidly improving generative AI, rather than any specific incident or particular detection method. Overall, the articles describe an ongoing cycle in which each new capability in deepfake generation pressures detection systems, while new detection approaches attempt to keep pace. The reports emphasize uncertainty and the need for continued development as AI-generated media becomes harder to verify through visual inspection alone.