Social media users are becoming increasingly concerned about content being falsely flagged as ‘adult’, and sometimes banned, amid an increase in some platforms’ use of AI moderation.
In recent months users have claimed that there has been an increase in such adult content flaggings on platforms including Instagram, Pinterest and Tumblr. It’s raised fears that AI moderation may not be trained well enough to identify genuine adult and explicit content on some platforms.
In May 2025 Pinterest apologized to users, following weeks of users reporting mass-bannings for supposed content violations, for content that did not breach guidelines. Among these were reportedly many false adult content flaggings.
Pinterest has said that it may use AI for “improving content moderation”, adding: “We use human reviewers and feedback from the public to refine our AI”.
The company told TechCrunch that the mass-bannings were not the result of AI moderation, and in a public apology blamed an unspecified “internal error” that led to “over-enforcement”. It added that accounts that had been mistakenly deactivated had been reinstated.
Pinterest added: “We’ve reinstated many impacted accounts and are making improvements to respond faster when mistakes happen going forward.”
Despite Pinterest’s AI denial, the seemingly mystifying mass-bannings coming in the age of AI moderation led to some users suspecting that human moderation on the platform wasn’t as thorough as it should be.
Complaints tumbl forth
Tumblr also came under similar suspicion, after a spate of the platform’s users recently complained that their content had been falsely flagged as “mature”.
Tumblr says it uses a mix of “machine-learning classification and human moderation”, adding: “Computers are better than humans at scaling process – and we need them for that – but they’re not as good at making nuanced contextual decisions.”
The false flag issue was reportedly linked to an update to the platform’s Android app, with Tumblr saying that its content preference settings moderation was “an ongoing process as we continue to fine-tune how we detect and address mature content”.
Tumblr also said that it was “actively working” to reduce “incorrect classification issues”, and had updated its appeals and reviews processes. The company added: “This is complicated and lengthy work, so we thank you for your continued patience.”
More Insta-bans
There also seems to have been a recent increase in the amount of Instagram users complaining about their accounts being banned or limited, despite claims that their content does not violate guidelines.
Again, this has led to speculation about increased use of AI in content moderation, although Meta has so far refused to comment publicly on this issue. An online petition called Meta Wrongfully Disabling Accounts with No Human Customer Support has been launched, attracting over 5,000 signatures so far.
The creators of the petition claimed that thousands of Instagram, Facebook and WhatsApp accounts had been wrongly disabled due to AI content moderation, although no direct evidence of this AI claim has been presented. SEXTECHGUIDE has contacted Meta and will update this article with any response.
Come back when you know what a wanger is
AI is, by its nature, technology that improves over time and with training, but it seems that some platforms may be handing too much moderation control to machine learning before it’s developed the nous to do so effectively.
The fact that many tech platforms have laboriously slow response times to complaints about account blockings, and sometimes don’t provide human replies at all, is hardly helping the issue.
AI content moderation is undoubtedly here to stay, but these systems clearly need to spend a bit more time studying tits and wangers before they’re allowed to take the lead in the judge, jury and executioner roles for online content moderation.
Leave a Reply