Watermarking ChatGPT, DALL-E and other generative AIs could help protect against fraud and misinformation

Digital watermarks on training data used for generative AI tools such as ChatGPT could help protect against fraudulent activity, an expert says.