AI and the New Reality of Sextortion: How Anyone Can Become a Target and How to Protect Yourself

Generative artificial intelligence has pushed sextortion to a dangerous new level. Where once victims were mainly adult industry workers or people who had shared intimate content with a malicious individual, today any social media user with an ordinary public photo can become a target. Recent discoveries of unsecured AI image databases show how quickly this threat is scaling and why everyone should understand the risks.

Sextortion, a blend of 'sex' and 'extortion', is online blackmail in which a criminal threatens to publish intimate images or videos unless the victim complies with demands or pays money. The threat relies on fear of reputational damage, family exposure, or professional harm, which is why victims often hesitate to come forward. The core mechanism has not changed, but the material used for blackmail is now far easier to fabricate.

Text-to-image and image-editing tools can now undress people shown in any digital photo or attach a generated nude body to someone's face within seconds. Faking such content once required skilled, time-consuming work in photo-editing software. Today, a user only needs to describe the desired result in words. Many generative AI services also show little concern for protecting the source images uploaded to them, meaning a victim's photo can be processed without consent and the output can be exposed. As a result, the pool of potential victims has expanded from a small high-risk group to almost anyone with a public online presence.

In October 2025, cybersecurity researcher Jeremiah Fowler found an open database containing more than one million AI-generated images and videos, the vast majority pornographic. The database was not encrypted or password-protected, so anyone on the internet could access it. Its name and watermarks on some images pointed to the American company SocialBook, which offers services for influencers and digital marketing and links to image-generation tools. Further analysis showed SocialBook did not directly create the content; its interface linked to third-party products, MagicEdit and DreamPal, which were the tools actually used. The database reportedly grew by about 10,000 files per day. SocialBook denied any connection to the database, and after the researcher notified the company, several pages referencing the two tools became unavailable.

MagicEdit and DreamPal were initially presented as tools for interactive visual and character experiences, but much of their functionality centered on sexual content. MagicEdit offered AI-driven clothing changes and stylizations that made photos of women more provocative, such as replacing everyday outfits with swimwear or lingerie. DreamPal, positioned as an AI chat service focused on role-play, openly targeted adults, invited users to create an ideal AI girlfriend, and stated that explicit content filters were disabled. Both services suspended operations, and their apps were removed from the App Store and Google Play. At the time of writing, DreamPal's site returned an error while MagicEdit appeared functional again.

Fowler reported two other open databases of AI-generated images earlier in 2025. One belonged to the South Korean service GenNomis and contained 95,000 entries, many showing people in a 'nude' state. It also included images depicting childlike versions of celebrities such as Ariana Grande, Beyoncé, and Kim Kardashian. These cases reveal a systemic problem: companies building AI image tools often neglect basic data security, and the outputs can include both non-consensual sexual imagery and material involving minors. Such incidents underscore that the harm is not limited to private messages or shared intimate content.

Women are disproportionately targeted, but men are not safe. A familiar scare tactic claiming a hacker has used a webcam to film someone visiting pornographic sites could become far more convincing when paired with AI-generated photos or videos. Anyone whose images are publicly visible, from students and professionals to parents and public figures, may be a potential target. Children and teenagers are especially vulnerable, and their photos shared online can be exploited to create inappropriate content.

Set all social media profiles to the highest privacy level possible, since public photos can serve as source material for AI-generated sexual content. Limit the personal data and images you share publicly, and regularly review who can see your posts. Reduce your digital footprint by removing old accounts and data where feasible, and be cautious about services that request access to your photos. Consider privacy-management resources that provide step-by-step guidance for securing accounts across platforms. Parents should talk with children about the risks of sharing photos online, monitor the social networks they use, and help configure privacy settings so images cannot easily be misused. Be wary of unexpected blackmail messages, avoid paying, preserve evidence, and report the incident to the platform and to the relevant authorities.

The risks of sextortion no longer depend on private messages or intimate exchanges. In the age of generative AI, ordinary published photos can be turned into compromising content. Protecting your privacy, controlling your public data, and teaching your family safe online habits are now essential steps for preserving reputation and peace of mind.