A class action lawsuit has been filed against streaming platform Twitch by one of its own streamers following the platform's recent changes regarding AI. Although one streamer on the platform is responsible for filing the Twitch lawsuit, it is a class action suit, allowing for all Twitch streamers to be represented.
Twitch has been one of the leading streaming platforms for years, but not without controversy. Everything from bad behavior by streamers to changes to the platform itself have made waves in recent history. Recent changes to the service involving AI training have also drawn the ire of viewers and streamers alike, which this lawsuit is now addressing.
On August 20, a class action lawsuit was filed against Twitch due to its recent implementation of utilizing Twitch stream data for AI training. The lawsuit was filed by Warren Pandiscia, a Twitch streamer. The platform gained negative publicity earlier in August when it revealed that streams could be used to train AI models belonging to Twitch parent company Amazon. However, this class action lawsuit claims that the AI training dates back to 2024, and that streamers are only now being made aware of the fact. The lawsuit alleges that this move is a breach of contract, and looks for recompense for streamers whose data has been used without notice or compensation.
Back in 2024, Mike Minton, who is currently the Chief Product Officer for Twitch, indicated that content streamed on the platform was being used for "prototyping" an AI model. Back then, Minton indicated that the use of streams as training data was "within the bounds of user trust" and privacy regulations. However, it seems that no official notice was made directly to streamers that their data would be used this way. As a result, up to two years' worth of streams may have already been utilized as AI training data. According to Twitch's current FAQ, that data may include clips, past broadcasts, pictures, text, and even Twitch streamers' chat.
چوونەژوورەوە
بۆ ئەوەی ئیدیعای شوێنی خۆت بکەیت لەسەر لیدەربۆرد!
چوونەژوورەوە
بۆ ئەوەی ئیدیعای شوێنی خۆت بکەیت لەسەر لیدەربۆرد!
Earler in August, Twitch announced that it was adding a setting that would let streamers to deny Amazon the ability to use their streams as AI training material. However, its inclusion led to an uproar among the streaming community, as streamers have to actually opt out. Many streamers indicated that they would have preferred that it be set off from the get-go. However, Minton made it clear that it's set that way intentionally, as he stated in a stream that if streamers had to actually opt in to allow their data to be used as training material, "nobody would opt in." Rather than accepting that a majority of its streamers would prefer their data not be used for such a purpose, it seems Twitch chose to work around it to do so anyway.
One issue the lawsuit also points out is that even those who now choose to opt-out of having their data used can't do anything about potential years' worth of back data having been utilized. The lawsuit claims that creators "will never be able to claw back" the intellectual property that Amazon may have already used, which it calls unlawful. Even then, if someone who has opted out appears on another channel in the chat or otherwise, they have no control over whether that second channel chooses to opt-out, so their data could end up being utilized anyway. So far, Amazon hasn't offered an official response to what has become just the latest in a string of AI-related lawsuits in recent years.
سەرچاوەی فەرمی بە زمانی ئینگلیزی: https://gamerant.com/twitch-ai-training-lawsuit/
