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.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.Security[#item_full_content]