Omanghana
Amazon and its livestreaming subsidiary Twitch are facing a proposed class-action lawsuit alleging that creators’ broadcasts, videos, images and chat messages were used to train commercial generative artificial intelligence models without their consent or compensation.
The complaint was filed in the United States District Court for the Northern District of California by streamer Warren Pandiscia, who is seeking to represent millions of Twitch creators potentially affected by the alleged practices.
According to the lawsuit, Amazon and Twitch collected extensive amounts of creator-generated material, including livestreams, archived broadcasts, video clips, photographs and chat logs, for use as artificial intelligence training data.
The plaintiff argues that creators uploaded content to Twitch with the understanding that it would be used to operate, promote and monetize the livestreaming platform. The lawsuit claims this did not amount to permission for Amazon to repurpose the material for separate commercial AI products.
The complaint includes breach-of-contract allegations, maintaining that Twitch’s earlier platform agreements provided a limited license tied principally to the operation and promotion of the service.
It argues that using creator content to develop artificial intelligence systems went beyond the purposes users could reasonably have expected when they agreed to those terms.
The legal action also points to subsequent revisions to Twitch’s Terms of Service that reportedly expanded the licensing rights available to Amazon and its affiliated companies.
According to the plaintiff, the revisions suggest that the previous terms did not clearly authorize Twitch to supply creators’ content to Amazon affiliates for artificial intelligence development. Amazon and Twitch may, however, argue that their licenses were already broad enough to cover data processing and related technological uses.
The dispute follows criticism of Twitch’s decision to introduce an AI-training preference that placed creators into the program by default. Users who did not want their content used were required to disable the setting manually through their privacy controls.
Many streamers argued that the platform should have requested affirmative permission before using their work for AI training, particularly when the resulting models could support commercial products.
The backlash intensified following remarks attributed to Twitch Chief Product Officer Mike Minton during a community question-and-answer session.
Explaining why the platform adopted an opt-out system instead of requiring users to opt in, Minton reportedly said, “If it was opt-in, nobody would opt in.”
Critics cited the comment as evidence that Twitch understood many creators would not voluntarily allow their content to be used for AI training. The lawsuit is expected to rely on the remark as part of its argument that meaningful consent was absent.
The complaint also challenges the practical effectiveness of Twitch’s opt-out mechanism. It alleges that the setting is not retroactive and may not prevent previously collected content from being used.
The restriction is also reportedly applied at the individual-channel level rather than across a creator’s entire presence on the platform. This means a creator who disables AI training for their own channel could still appear in another streamer’s broadcast.
Their voice, face, screen name or chat messages could therefore potentially be captured through content uploaded by another user, despite their decision to opt out on their personal channel.
These limitations are particularly significant on Twitch, where creators frequently collaborate, participate in multiplayer broadcasts and communicate through other channels’ live chats.
The case raises broader questions about whether technology companies can rely on platform licenses to use historical user content for purposes that were not widely contemplated when the material was uploaded.
It also adds to a growing number of legal disputes involving authors, artists, publishers, photographers and other creators who allege that their work was used to train generative AI systems without permission or payment.
The plaintiffs are expected to seek damages, restitution and possible restrictions on the continued use of creator content. However, the allegations have not yet been proven, and the court must first determine whether the case qualifies to proceed as a class action.
Amazon and Twitch will have an opportunity to challenge the claims, the proposed class definition and the plaintiff’s interpretation of the platform’s historical terms.
The outcome could influence how streaming and social media platforms disclose artificial intelligence training practices, structure user licenses and obtain consent before repurposing creator content for new commercial technologies.
Source: Omanghana
