<ul><li>A study by CSIRO and SKKU found that none of the 16 leading deepfake detectors tested could consistently detect real-world deepfakes.</li><li>Deepfake detection tools are facing major weaknesses, making it easier for AI-generated fake images, videos, and audio to spread false information.</li><li>Current detectors fail due to limited scope and specific types of fake content they can effectively detect, emphasizing the need for more advanced strategies.</li><li>Researchers suggest using multi-method detection systems, including fingerprinting techniques and larger, more diverse datasets, to improve deepfake detection.</li></ul>

Scientists warn of major weaknesses in deepfake detection tools

Discover more