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