More than 24 million people visited AI-powered 'nudify' websites in September, according to a new analysis by Graphika. These apps use deep-learning algorithms to digitally alter images, primarily of women, making them appear naked without consent. The report highlights a 2,000% increase in spam ads directing users to such services since early 2023.
The apps are promoted on major platforms including YouTube, Reddit, and X, with 52 Telegram groups identified as channels for accessing non-consensual intimate imagery (NCII). A Telegram spokesperson stated the platform actively moderates harmful content and removes breaches of its terms of service.
Incidents have been reported in schools: a New Jersey high school saw AI-generated nude images of female students circulated, prompting a mother and her 14-year-old daughter to advocate for better protections. Similar cases occurred in Seattle and southern Spain, where the 'Clothoff' app was used to target over 20 girls.
Graphika's report warns that the increasing accessibility of these services will likely lead to more online harm, including harassment, sextortion, and child sexual abuse material. The apps operate on a 'freemium model', charging users from $1.99 to $299 for enhanced features.
Advertisements for these apps are often overt, using terms like 'undressing', while others disguise themselves as 'AI art services'. YouTube tutorials showing how to use 'nudify' apps remain available. The US Federal Trade Commission has raised concerns about generative AI's impact on consumer protection, though no federal laws currently ban deepfake pornography.
