Making Pornhub Safe: Tech Solutions to End Child Exploitation | Soloman Friedman & Sarah Bain ECP

Sure, well, why don't we talk about it. It's one thing to be able to say our platform is safe, our platform is among the safest online platforms in the world — that doesn't mean much if you can't back it up with the reality. So let's talk about that. I'll start with the uploader side, because a major concern for anyone who cares about the safety of the vulnerable is ensuring that all content posted to the platform is made by consenting adults who are consenting to the acts depicted, to the recording of it, and to the sharing of it. We are right now the only free video sharing platform on the planet that requires age ID and consent verification for every single person depicted in uploaded content. It is impossible to upload content — we're not YouTube, we're not Vimeo, we're not Facebook. You can't just create an account and start uploading. In order to upload, you need to be a verified uploader. That means we've verified that your government ID is legitimate, that your face matches the government ID, and that you are a real person by submitting to what we call a liveness scan, where you move your face around to ensure you're not just a screen capture. That is then verified by a moderator, an actual person. So first AI and then a moderator. We're big believers — and I know this is contrary to the trend right now as we see big social media companies cutting their trust and safety staffs and budgets — we're of the view that humans are required for this. We use AI to support humans. Our company Aylo is a company of about 1,500 people — Aylo is the parent company of Pornhub and other platforms — a full 20% of that workforce works exclusively in trust and safety. That's a massive investment. We have moderators who work 24/7 in shifts to moderate content and verify identity. Once we've verified identity and you're given permission to post, nothing goes live immediately. Every single piece of content is scanned first by about 12 AI tools looking for matches with known child sexual abuse material from databases from organizations all around the world, including in the United States. We subscribe to any database we possibly can, including our own in-house tools. But every piece of content is also reviewed by a moderator ensuring that the identity of the people matches the identity on file, and that the acts depicted — and this is going to come as a big surprise — most people think as long as it's adult, anything goes, and it couldn't be further from the truth. There are many things we won't allow even though they're totally fictional and consensual and it's just acting. We don't allow depictions of serious violence that appear likely to cause bodily injury or death. We don't allow people to pretend to be drunk, to pretend to be blackmailing each other, or to pretend to be engaging in non-consensual activity even if it's totally consensual. Our platform is just not a place for that. We don't allow role-playing as minors or as real relatives. These are many safeguards all done in the moderation process, and it costs a lot of money — it's expensive, but to us it's an investment that's well worth it because we have a responsibility to keep the platform safe. One of our roles as investors is to support that investment and ensure that we're actually enhancing those capabilities and that workforce. It took a lot to get there. If you had asked a performer five years ago to give their real ID to a porn website, they would have looked at you like you were crazy. It just wasn't done. We had to build that trust with our community. I am so open to criticism — I want to hear feedback. But it's one thing to sling mud anonymously on Twitter, and another thing to sit down and collaborate with us. I go to law enforcement conferences. I went to the largest counter-internet child exploitation conference in Canada. I spoke to an audience of 400 police officers and prosecutors. I want to hear from the experts as to where the issues are. By the way, that audience was blown away by what we're doing in terms of trust and safety. I just scratched the surface in terms of how we keep people safe, but we want to engage proactively, and you only do that with dialogue.