US Edition
Canadian researchers create tool to remove anti-deepfake watermarks from AI content
CANADA, JUL 23 – The UnMarker tool removes AI-generated content watermarks over 50% of the time, revealing weaknesses in current watermarking methods aimed at combating misinformation.
- Canadian researchers at the University of Waterloo developed UnMarker, a tool that removes AI image watermarks even without knowledge of the watermarking method.
- This research emerged as AI companies race to implement watermarking to identify AI content, but watermarks have known vulnerabilities and face growing threats from misuse.
- UnMarker is the first universal practical tool that destroys watermarks in real-world settings within two minutes without interacting with watermark detectors.
- Kassis noted that watermarks are widely regarded as the primary method for defending against AI-generated content, but his research reveals a fundamental weakness that allows attackers to remove them.
- The findings suggest that current reliance on watermarking is insufficient, and Kassis calls for broader efforts to develop alternative solutions against deepfakes and AI content misuse.
25 Articles
25 Articles
UW researchers figured out how to bypass anti-deepfake markers on AI images
Researchers at the University of Waterloo have created a program that can recreate deepfaked images without the watermarks that show it's AI origins, exposing a systemic vulnerability in the defense against deepfakes, according to researchers.
Canadian researchers build tool to remove anti-deepfake watermarks from AI content
University of Waterloo researchers have built a tool that can quickly remove watermarks identifying content as artificially generated — and they say it proves that global efforts to combat deepfakes are most likely on the wrong track.
Watermarks offer no defense against deepfakes, study suggests
New research from the University of Waterloo's Cybersecurity and Privacy Institute demonstrates that any artificial intelligence (AI) image watermark can be removed, without the attacker needing to know the design of the watermark, or even whether an image is watermarked to begin with.
Winnipeg Free Press
Canadian researchers create tool to remove anti-deepfake watermarks from AI content
Breaking News, Sports, Manitoba, Canada
It would prove that the fight against hyperstructure is probably on the wrong track. The post A new tool removes the watermarks identifying the AI appeared first on Les Affaires.
Coverage Details
Bias Distribution
- 75% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium
