AI Deepfakes of Late-Night Hosts Like Jimmy Kimmel Spread on Social Media

Late-night TV hosts like Jimmy Kimmel are increasingly becoming targets of AI-generated deepfake videos spreading across social media platforms, raising concerns about misinformation and free speech. These fabricated clips, often created using generative AI technology, mimic the hosts' appearances and voices to spread misleading content, sometimes garnering hundreds of thousands of views before being identified as fake.

Analysts warn that late-night hosts are particularly vulnerable to deepfakes due to their frequent political commentary and widespread online presence. Jed Rosenzweig, founder of LateNighter.com, noted that audiences may unknowingly trust these AI-generated videos because they resemble authentic late-night show segments. "Jimmy Kimmel really does go on television and talk about politics all the time," Rosenzweig said. "So if you see a Kimmel thumbnail tied to something Trump did yesterday, there's a built-in context and trust there."

The ease of creating these deepfakes stems from the hosts' static presentation—often sitting or standing in one place—and the availability of high-quality audio recordings of their voices. Hany Farid, a computer science professor at Dartmouth College specializing in deepfake analysis, explained that AI tools can animate a single image with cloned voice and script. "Here's a voice, fifteen seconds. Here's a script. Go," Farid said.

These AI-generated monologues pose a unique threat because late-night TV remains a key source of news and opinion for many Americans, particularly Democrats who consume more clips online than full shows. Farid added that deepfakes leveraging trusted voices are particularly effective disinformation tools. "It's essentially disinformation campaigns that now seem to be coming out of the voices of those very people that we're used to listening to," he said.

While some deepfakes aim to spread political messages, others appear to be financially motivated clickbait. Farid suggested creators may be driven by ideology or revenue, noting that platforms like YouTube and TikTok often host these videos long enough to accrue significant engagement before moderation takes effect. A Jon Stewart deepfake about Italian Prime Minister Giorgia Meloni remained on TikTok for over two months without an AI label, while a fabricated Jimmy Kimmel video discussing former President Donald Trump’s golf game amassed 230,000 views on YouTube.

Social media companies have implemented policies to address deepfakes, requiring creators to label AI-generated content and using detection tools to identify unauthorized simulations. YouTube, TikTok, and Meta—parent company of Instagram and Facebook—stated they remove harmful deepfakes and require disclosure for realistic AI content. However, the volume of content makes enforcement difficult, and legal protections remain limited.

Louis Tompros, an intellectual property litigator specializing in AI, noted that current copyright law does not adequately address unauthorized synthetic doubles of public figures. While some states like California and Tennessee offer protections against AI voice clones and deepfakes, broader federal solutions are still under development.

As generative AI tools become more accessible, the spread of fake late-night host videos highlights the ongoing challenge of combating digital misinformation and protecting public trust in media.

#Deepfakes #JimmyKimmelLive #LateNightTV #AIinMedia #SocialMediaMisinformation #GenerativeAI #FreeSpeech