Pornhub AI chatbot intercepts millions hunting for child abuse videos

Key Points:

  • The pop-up chatbot on Pornhub intercepted millions of people searching for child abuse videos
  • The chatbot resulted in a decrease in the number of searches for material about child sexual abuse
  • A significant number of users were directed to charitable help services for support


Pornhub’s UK website has deployed a chatbot that has thwarted over four million individuals searching for child abuse material on the platform in the past two years. The chatbot, triggered by specific words or phrases associated with abuse, issues a pop-up alerting users that the content is illicit, ultimately guiding them towards assistance services in 50% of cases.


Conducted in collaboration with two UK child protection organizations, this innovative approach has been found to significantly reduce searches for material related to child sexual abuse. Joel Scanlan, senior lecturer at the University of Tasmania who spearheaded the reThink Chatbot evaluation, emphasizes the effectiveness of the deterrence messages in curbing such searches.


During the trial period, a staggering 4,400,960 warnings were instigated by searches on Pornhub’s UK website, with the vast majority—99%—unaffected by the chatbot. The interactive chatbot engages users triggering the pop-up with queries and educational prompts, urging them towards charitable resources for support.


Pornhub, in light of prior scrutiny over child exploitation and nonconsensual content on the platform, has intensified efforts to combat illegal material. The site annually removes millions of images and videos associated with child sexual abuse, employing a stringent list of 34,000 banned terms to monitor and prevent such material as per a company spokesperson cited by WIRED.


In broader regulatory contexts, the EU has implemented stricter measures addressing adult content platforms, while some U.S. states like North Carolina and Montana have restricted access to sites like Pornhub following the enactment of age verification laws earlier this year.


The collaborative report further reveals that the chatbot prompted 1,656 requests for additional information and directed 490 individuals to the charity’s Stop It Now website. Notably, around 68 individuals sought assistance through Lucy Faithfull’s confidential helpline, underscoring the tangible impact of this innovative intervention.



Prompt Engineering Guides



©2024 The Horizon