Twitch is, Vaguely, Working to Fix its Child Predator Problem

Twitch has been hounded for months after a September report showed just how many sexual predators were stalking the streaming platform’s halls, targeting children that were not supposed to be there in the first place. Now the company says it’s developed systems to combat child sexual abuse material, though it’s not exactly sharing the details.

In a Tuesday blog post, Twitch said it is working on a “constantly evolving approach” to limit harm to young people. The user-streaming platform said it’s creating phone verification requirements for “potentially vulnerable accounts,” AKA those accounts made by young people pretending to be over 13, before they can livestream. The company said it’s also working to delete more accounts for those under 13-years-old.

In addition, the Amazon-owned platform has blocked certain search terms in the in-app browser. The company seemed to be purposefully vague about which search terms it has blocked and how its phone verification system would work. The platform mentioned it is working on creating a machine learning system to detect harm in written text on Twitch, but the company offered no specifics about its planned development or implementation.

Gizmodo reached out to Twitch to see if the company would expound on how its phone verification system would work, though a company spokesperson responded Twitch wanted to be purposefully vague.

“We always need to be careful with the amount of specificity we offer you [referring to Twitch users and the public] in order to avoid giving bad actors information they could use to evade our efforts—and this is especially true when it comes to child predators,” the company wrote on its blog. “There is no single fix to prevent predation,”

This comes just a few months after a massive report from Bloomberg showed child predators were using Twitch’s services to track and interact with young children streaming on the platform. Bloomberg’s data, based on an anonymous internet extremism researcher, showed that some predatory accounts were targeting hundreds of thousands of young people’s accounts. The researcher found streams of young people and tracked which accounts were also following other children. There were hundreds of these kinds of predatory accounts following more than 1,000 children each.

The company mentioned that it has updated its privacy settings on the internal messaging whispers system, which in testing revealed that blocking strangers from messaging new users is turned on by default. Twitch also claimed it will prioritize user reports that involve young users.

Twitch’s user policy says account-holders must be 13 years or older, and there’s the usual slate of banned violent or sexual content, but just like many other digital platforms, there’s been a struggle to keep child sex abuse material from appearing on the site. Twitter has a problem with this, too, which has led to advertisers abandoning the site. The National Center for Missing and Exploited Children’s 2021 report on instances of child sexual abuse material showed Twitch was ranked rather low compared to other services, with 6,629 reports compared to TikTok’s 154,618 or WhatsApp’s 1.3 million. Twitch mentioned its law enforcement response team shares information with police and the NCMEC.

Still, Twitch said it wanted to crack down even more on grooming on the platform, which involves adults interacting with young people before manipulating them and eventually exploiting or abusing them. It’s hard to tell without specifics, but Bloomberg’s report mentions Twitch’s existing tools have previously proved incapable of stopping adults from contacting and, in some cases, grooming young people. Twitch Chief Product Officer Tom Verrilli told Blomberg that “even one single instance of grooming is abhorrent to us.”