Runway, a company developing tools for generating videos based on artificial intelligence, has found itself at the center of a scandal. 404 Media published investigation alleging that Runway used thousands of YouTube videos, including content from well-known media companies and popular bloggers, to train its AI model without permission.
Po message The Verge, 404 Media gained access to table with data that Runway allegedly used to train its AI. It contains links to YouTube channels from companies like Netflix, Disney, Nintendo, and Rockstar Games, as well as popular bloggers like MKBHD, Linus Tech Tips, and Sam Kolder. The spreadsheet also contains links to news outlets including The Verge, The New Yorker, Reuters, and Wired.
According to a former Runway employee, the spreadsheet consists of a list of channels that are “the result of a company-wide effort to find quality videos to build a model.” These links were then used as input to a web crawler that downloaded the content using proxy servers to avoid Google blocking.
In addition to YouTube channels, the data found includes links to pirated sites that show licensed movies for free. It is not yet clear whether Runway used these movies to train its Gen-3 Alpha model, but as The Verge explains, it will likely be impossible to know. The publication asked Runway to comment on the situation, but has not yet received a response.
It’s worth noting that Runway isn’t the only company using YouTube to train its AI models. Earlier this year, OpenAI CTO Mira Murati said she wasn’t sure whether its AI model for generating the full HD Sora video was trained on YouTube content. Additionally, a recent joint investigation by Proof News and Wired found that Anthropic, Apple, Nvidia, and Salesforce trained their AI models on more than 170,000 YouTube videos.
If you notice an error, select it with your mouse and press CTRL+ENTER.
