Critical flaws have been uncovered in widely used deepfake detection tools, sparking a call for urgent improvements in deepfake detection technologies.
A study by Australia’s CSIRO and Sungkyunkwan University in South Korea, assessed 16 leading detectors and found none could reliably identify real-world deepfakes.
CSIRO cybersecurity expert Sharif Abuadbba said deepfakes were artificial intelligence (AI) generated synthetic media that could manipulate images, videos, or audio to create hyper-realistic, but false content.
Dr Abuadbba said the availability of generative AI had fuelled the rapid rise in deepfakes, which were cheaper and easier to create than ever before, raising concerns about misinformation, fraud, and privacy violations.
“Deepfakes are increasingly deceptive and capable of spreading misinformation, so there is an urgent need for more adaptable and resilient solutions to detect them,” he said.
“As deepfakes grow more convincing, detection must focus on meaning and context rather than appearance alone.”
Dr Abuadbba said breaking down detection methods into their fundamental components and subjecting them to rigorous testing with real-world deepfakes would enable the development of tools better equipped to counter a range of scenarios.
He said the study found many current detectors struggle when faced with deepfakes that fell outside their training data.
“For example, the ICT (Identity Consistent Transformer) detector, which was trained on celebrity faces, was significantly less effective at detecting deepfakes featuring non-celebrities.”
Fellow CSIRO cybersecurity expert Kristen Moore said using multiple detectors and diverse data sources strengthened deepfake detection.
“We’re developing detection models that integrate audio, text, images, and metadata for more reliable results,” Dr Moore said.
“Proactive strategies, such as fingerprinting techniques that track deepfake origins, enhance detection and mitigation efforts.
“To keep pace with evolving deepfakes, detection models should also look to incorporate diverse datasets, synthetic data, and contextual analysis, moving beyond just images or audio.”
Read the full paper: SoK: Systematization and Benchmarking of Deepfake Detectors in a Unified Framework.
