A new study has raised questions about the role of the world’s biggest social media networks. YouTube and X (now Twitter) are steering millions of people to AI-based “nudify” apps.
The apps make fake nude images of the subjects from ordinary photos, and they are more available to users even as privacy and harassment are a concern.
As a result, users looking for AI image editing tools or related keywords are often exposed to content promoting nudify services. Recommendation algorithms, promotional posts and advertisements on popular platforms are helping these applications reach a wider audience, including minors, the study says.
Nudify apps lean on AI image generation models to digitally manipulate photographs by creating fake explicit images. While many developers market such tools as entertainment or novelty software, experts say they can be easily used for harassment, blackmail, cyberbullying, revenge pornography, and non-consensual image creation.
Promotional videos, influencer content, and sponsored posts are driving high traffic to these services, the research showed. Some advertisements have been found to bypass platform moderation by using coded language, emojis or alternative spellings to hide their features from detection. When people click on these links, they are often directed to third-party sites e.g., subscription or AI-based services and offered subscriptions or premium tools.
Digital rights advocates say the fast spread of AI-generated explicit imagery shows the increasing difficulty in regulating generative AI technologies. People who are victims of AI-generated fake images suffer emotional distress, reputational damage and legal challenges, especially when manipulated images spread widely on social media.
Technology companies have repeatedly said they prohibit content promoting sexually explicit deepfakes and non-consensual intimate imagery. Both YouTube and X have policies to remove harmful content and limit the spread of abusive material. But researchers say enforcement remains inconsistent and promotional content and links are still visible for long periods.
The study also illustrates the larger problem of balancing free expression and user safety. With more sophisticated and accessible generative AI tools, platforms are more under pressure to improve automated detection systems, enhance moderation, and respond more quickly to reports of harmful content.
Others are mulling tougher laws on AI-generated deepfakes. More than a dozen countries have passed or are considering laws that would prohibit the creation or distribution of non-consensual intimate images, whether they are real or AI-generated. Privacy experts believe stronger legal frameworks as well as better platform oversight are needed to stop the increasing misuse of generative AI.
To share personal photographs online, cybersecurity professionals say people should be careful when sharing their own photos online and stay away from any suspicious AI image generation sites. They also urge people to report harmful content and verify the legitimacy of AI tools before using them.
With artificial intelligence changing how digital content is created and developed, the findings of this study have increased calls for greater transparency, more accountability on the part of platforms, and stronger controls to prevent AI-enabled abuse while allowing responsible innovation to continue.
