Meta's Screening Failure: Sexual Content Involving Minors Approved for Publication
Meta has removed dozens of advertisements from its network containing offensive sexual content involving minors, generated by AI, after researchers from the Tech Transparency Project identified more than 50 such ads. The company announced it is developing new systems to strengthen enforcement.
Researchers from the Tech Transparency Project (TTP), an independent American organization that monitors the activities of technology companies, identified dozens of paid advertisements that included offensive sexual material created using artificial intelligence featuring minors. The American technology magazine WIRED verified the findings with Meta and published the investigation.
According to the investigation, between November 2025 and early August 2026, fake sexual images appeared in more than 50 ads on Facebook, Instagram, Threads, and Messenger. The ads were identified in the USA, the UK, and more than a dozen countries in Europe. Some of them promoted applications called "Nudify," which use artificial intelligence to create fake sexual images of people without their consent.
According to TTP director Katie Paul, "the ads made no effort to hide the images or what they were promoting." The researchers noted that some of the campaigns directed users to services like MaskAI, an application that they say was later removed from Apple's App Store after it violated the company's policy prohibiting digital "stripping" applications.
According to TTP's findings, some of the ads combined images of minors with messages of an explicitly sexual nature. In one documented case, a video ad displayed a thumbnail of a child alongside text inviting users to "fulfill fantasies" using artificial intelligence. After clicking on the ad, sexual videos were displayed in which the minor's face was artificially inserted, according to the researchers.
Researchers stated that even after some of the campaigns ceased to be active, the ads remained available for months in Meta's public ad library, which indicates that they passed the company's approval mechanism. In one of the campaigns, 2,563 impressions were recorded for users in Europe.
According to WIRED magazine, Meta's ad review system relied heavily on automated systems based on pattern recognition, which struggled to identify the context of the content. According to the researchers, this is why ads were approved that apparently violated the company's policy, which prohibits the exploitation or harm of children as well as any explicit sexual content in the advertising system.
In response, Meta stated that it removed the ads and accounts that were reported to it. The company further stated that it is developing a new generation of artificial intelligence models designed to improve enforcement against content involving the sexual exploitation of children. According to the company, the new systems examine the full context of posts and ads — including language, intent, and wording — and do not rely solely on pattern recognition.
The company added that in sensitive cases, "human reviewers continue to be involved."
Another legal blow: A huge fine of nearly a billion dollars
At the same time, the exposure of the ad failure comes as Meta suffers an unprecedented legal blow regarding the issue of safety and protection of children on its platforms. A court in the state of New Mexico in the USA ordered the company to pay an additional fine of 567 million dollars — this is in addition to a fine of 375 million dollars imposed on it in March last year due to the damage of social networks and the addiction of minors. Thus, the total amount of compensation that the tech giant must pay in the state rises to 942 million dollars.
Alongside the heavy financial sanction, the judge issued an unprecedented order requiring Meta to make deep technological changes in the way Facebook and Instagram operate for young users in its jurisdiction. Among other things, it was determined that the company will have to remove the "like" counter for minors under the age of 18, unless explicit permission is received from a parent or guardian.
In addition, there is an obligation to disable notifications for young users at night (between 10:00 PM and 07:00 AM), and to limit their total usage time on the networks to up to 90 hours per month — an average of about 45 minutes per day.
"A significant number of people in New Mexico are experiencing harm from Meta's products due to risks of sexual exploitation, interference with education, and negative mental health consequences," the court's decision stated. New Mexico Attorney General Raúl Torrez sharply attacked the company: "For years, Meta knew that its platforms were harming children — from fueling a mental health crisis to connecting sex offenders with children — and chose engagement and profits over their safety. Today, Meta is paying for that choice."
On the other hand, Meta clarifies that it intends to appeal the verdict. Company spokesperson Andy Stone stated that Meta is working hard to keep users safe.
