Tom Hanks and MrBeast Warn About AI Deepfake Scams

This week, actor Tom Hanks, CBS Mornings host Gayle King, and popular YouTuber MrBeast warned their followers that scammers are using their likenesses in advertisements created with artificial intelligence (AI).

Tom Hanks Alerts Fans to Deepfake Dental Ad

Tom Hanks was the first to notify his followers, posting on Instagram that his image was being used in an AI-generated ad for a dental plan scam. The actor emphasized that he has no connection to the product, stating, “I have never heard of this product and have never used it! Please don’t be fooled by these AI videos.”

Gayle King and MrBeast Also Targeted

Soon after, CBS host Gayle King issued a similar warning after discovering her likeness was being used in a deepfake ad for weight loss products.

Popular YouTuber MrBeast also addressed his followers on X (formerly Twitter), saying he was shown a deepfake video where his speech and lip movements were generated by AI. In the video, scammers promoted a fake iPhone 15 giveaway. MrBeast shared the fake video in his post, which can be seen below.

“A lot of people have seen this scammy deepfake ad with me in it,” MrBeast wrote. “Are social media platforms ready to handle the rise of deepfakes and AI usage? This is a serious problem.”

How AI Tools Are Used in Scams

It appears that all of these scam ads were created using AI tools that can alter a person’s facial movements and synthesize voices based on samples freely available online. Experts have repeatedly warned that convincing deepfakes created with AI are becoming a growing problem, threatening trust and the reliability of modern communication technologies, as people may soon have to doubt everything they see and hear.

Efforts to Combat Deepfakes

Currently, tech giants like Google and OpenAI are working on implementing “watermarks” and special metadata in AI-generated content to help track its origin. However, researchers have already demonstrated that such watermarks can be easily bypassed. Additionally, open-source AI tools exist—and will continue to exist—that do not mark content with any watermarks at all.