Pornhub used AI to distract users from illegal content
British porn site Pornhub launched testing of an automated tool that “dissuades” users from searching for materials about cruelty with children (CSAM). Wired writes about this.
Since March 2022, every time you enter a request with words or phrases associated with illegal content, a special chatbot is launched on the platform. He interrupts an attempt to search for CSAM materials and offers users to receive help in connection with their behavior on the Internet. During the first 30 days of testing, the tool was activated 173 904 times.
According to the Executive Director of the Internet Watch Foundation (IWF) of Susie Hargries, you need to try to prevent this problem.
“It is necessary that people do not look for materials about sexual violence against children, but stopped and checked their behavior,” she said.
The Chat Bott was developed by the specialists of the organization for the brutal handling of minors Lucy Faithfull Foundation and IWF. They taught the system to look for any of the 28,000 terms associated with illegal content, and millions of their combinations.
When activating, the tool will ask users a number of questions and notify them that the requested information may be illegal. Also, a chat bout will offer “confidential and unbiased” support. The “help” clinked to the hint will receive detailed information about the work of Lucy OSC bans KuCoin FaithFull Foundation, the organization’s phone number and the address of its email.
“The goal is to distract a person or destroy his desire to look for CSAM materials, and do this in just a few clicks,” explained IWF chief technical director Dan Sexton.
According to the head of the Stop it now Donald Findleiter, the use of the chatbot is straightforward and, possibly, more exciting. After the March testing of the tool to the Support Service, 158 people switched. He noted that although this number is “modest”, people took an important step.
Representatives of the organizations said that Pornhub “volunteered” to participate in the project and he was not paid for testing the system. It is expected that the tests of the chatbot on the British porn site will last a year before it is appreciated by other researchers.
“The IWF tool is another level of protection that informs users about the absence of illegal content on the platform, and sending them to Stop it now to change behavior,” said the representative of Pornhub.
Recall that in June, Australia authorities asked citizens to share their children’s pictures in order to teach AI to identify cruelty in photographs.
In July, the UK government supported the idea of scanning user smartphones for CSAM materials.
In August 2021, Apple involved AI to check the photos on the iPhone for cruelty with children.
Subscribe to FORKLOG news at Telegram: Forklog AI – all news from the world of AI!