Teenagers in the US were able to access 135 artificial intelligence apps capable of creating nude or sexually explicit images despite age restrictions and parental controls, according to an investigation by the Tech Transparency Project (TTP).
The findings raise questions about how effectively Apple and Google screen software available through their app stores, particularly when the technology can be used to target classmates without their consent.
Some of the applications identified during the tests carried age ratings that allowed children to access them, even when restrictions were enabled. The investigation also found that searches for seemingly ordinary editing tools could lead to applications with sexualised features.
The findings come as schools across the US deal with cases involving AI-generated intimate images of students. Apple has since strengthened its app review rules, while Google says it has suspended offending applications and restricted related search terms.
AI Nudify Apps Bypassed Apple and Google Age Restrictions
The investigation, published on 19 October by TTP, used four test accounts registered as belonging to 13-year-olds. Researchers created two Apple accounts and two Google accounts, with parental controls activated on one account for each platform.
They searched for applications using phrases such as 'AI video generator' and 'adult AI', rather than relying exclusively on obvious terms associated with nudification. This approach uncovered software that could digitally remove clothing or place a person's face onto a nude body.
According to the report, one face-swapping app capable of generating nude images carried an age rating suitable for children aged four and above. Separately, the report described another app rated for the same age group whose advertising showed sexually explicit imagery.
The free features of one app allowed users to place women's faces on lingerie-clad bodies, while paid options offered more explicit scenarios.
The researchers also found that subscriptions or in-app credits could be purchased through a 13-year-old's account without an alert reaching the linked parental account. The purchases were made using gift cards. Apple said parental approval should have been requested and that it was investigating why this had not happened.
The scale of the app ecosystem is another concern. It was further reported that the applications identified in its investigation had collectively recorded 5.5 billion downloads and generated $3.9 billion (£2.9 billion) in revenue. These figures cover the apps identified by the organisation, rather than downloads or earnings attributable solely to underage users.
Researchers used AI-generated images of fictional women during their testing, rather than photographs of real people. The investigation therefore examined the applications' capabilities without creating synthetic intimate images of actual individuals as part of the tests.
Apple and Google Respond To AI Deepfake Concerns
The findings have real-world relevance for families dealing with the misuse of photographs. In Pennsylvania, Morgan Dorfman discovered in December 2025 that a student at her daughter's high school had used the Movely editing app to create sexual videos featuring her daughter and five other girls.
Dorfman later investigated the software herself to understand how it worked and help other parents recognise the risks. She found that its free version promoted suggestive video templates, while paid features offered more explicit material.
The case illustrates how image-editing tools can be used to create sexualised content featuring people who have not consented. Those affected may experience distress at school and face uncertainty about how to report the material or limit its circulation.
Google said it takes 'proactive steps to detect and remove apps with harmful content'. Spokesperson Dan Jackson added that the company investigated reported violations, suspended offending apps and restricted related search terms.
Apple also defended its review process. Spokesperson Adam Dema said some applications identified in the report were mainstream editing tools that did not possess nudification or deepfake capabilities. He said the company prohibited nudify apps and removed developers who introduced banned features after approval, a practice known as 'bait-and-switch'.
Following contact from TTP, Apple updated its App Review Guidelines to require developers to prevent and remove non-consensual intimate imagery, including AI-generated material. The company also said it had taken action against applications identified in the investigation.
However, removing an app from a store does not necessarily disable a copy already installed on a device. TTP reported that applications removed during its tests remained functional on the test phones, leaving a potential gap between store enforcement and preventing continued use.
The findings do not establish that every teenager can access all the applications identified, nor do they mean parental controls are ineffective in every situation. Instead, the investigation highlights the difficulty of relying on age ratings when an application's features may not match its classification.
For parents, the results underline the limits of relying on app-store restrictions alone. For the companies, the challenge is to detect prohibited capabilities during review and respond when apps introduce hidden features or violate their policies after approval.
