A group of YouTubers suing the tech giant for scraping their videos without permission to train AI models has added Snap to its list of defendants. The plaintiffs, an internet content creator who operates three YouTube channels with a total of about 6.2 million subscribers, allege that Snap trains its AI systems on its video content and features AI features such as the app’s “Imagine Lens,” which allows users to edit images using prompts.
Plaintiffs previously filed similar lawsuits against Nvidia, Meta, and ByteDance over similar issues.
In a new proposed class action lawsuit filed Friday in California’s Central District Court, YouTubers specifically accuse Snap of using a large video language dataset known as HD-VILA-100M, as well as other datasets similarly designed solely for academic research purposes. Plaintiffs allege that by using these datasets, Snap circumvented YouTube’s technical restrictions, terms of service, and commercial license restrictions.
The lawsuit seeks statutory damages and a permanent injunction against future claims of copyright infringement.
The lawsuit itself is being led by the creators of the h3h3 YouTube channel, which has 5.52 million subscribers, and smaller golf channels MrShortGame Golf and Golfoholics.
This is one of many competitions currently pitting content creators and AI model providers against each other, including those focused on copyright disputes by publishers, authors, newspapers, user-generated content sites, artists, and others. This is not the first case involving former YouTubers. More than 70 copyright infringement lawsuits have been filed against AI companies, according to the nonprofit Copyright Alliance.
In some cases, judges have ruled in favor of the tech giants, such as the case between Mr. Mehta and a group of authors. In some cases, like the case between Anthropic and a group of authors, the AI giant has paid compensation to plaintiffs for their claims. Many lawsuits are still pending.
tech crunch event
san francisco
|
October 13-15, 2026
Snap has been contacted for comment. TechCrunch will update if available.
