Facial recognition and machine learning programs have officially been democratized, and of course the internet is using the tech to make porn. As first reported by Motherboard, people are now creating AI-assisted face-swap porn, often featuring a celebrity’s face mapped onto a porn star’s body, like Gal Gadot’s likeness in a clip where she’s supposedly sleeping with her stepbrother.

But while stopping these so-called deepfakes has challenged Reddit, Pornhub, and other communities, GIF-hosting company Gfycat thinks it’s found a better answer. While most platforms that police deepfakes rely on keyword banning and users manually flagging content, Gfycat says it’s figured out a way to train an artificial intelligence to spot fraudulent videos.

The technology builds on a number of tools Gfycat already used to index the GIFs on its platform. And the new tech demonstrates how technology platforms might try to fight against fake visual content in the future.

That battle will likely become increasingly important as platforms like Snapchat aim to bring crowdsourced video to journalism. Gfycat, which has at least 200 million active daily users, also hopes to bring a more comprehensive approach to kicking deepfakes off a platform than what Reddit, Pornhub, and Discord have managed so far.

Mashable reported on Monday that Pornhub had failed to remove a number of deepfake videos from its site, including some with millions of views. (The videos were later deleted after the article was published). Read more from wired.com…

thumbnail courtesy of wired.com