Amazon and its subsidiary, Twitch, are facing a class-action lawsuit regarding the use of livestreamed content to train artificial intelligence models. The legal action, initiated by Connecticut-based streamer Warren Pandiscia, claims that the company exploited the work of millions of creators without obtaining proper authorization or providing financial remuneration.
The complaint argues that the unauthorized use of this data constitutes a breach of contract between the platform and its users. The lawsuit seeks both monetary damages for affected streamers and a court order to halt the practice of using such content for AI development.
Controversy surrounding the data usage intensified in August when Twitch introduced an opt-out mechanism for users. However, the effectiveness of this feature has been questioned, as it only applies to individual streams and does not prevent a creator's content from being captured if they appear in a broadcast where the setting remains active. Furthermore, it remains unclear when Amazon began collecting this data. Mike Minton, Twitch’s chief product officer, previously stated he was uncertain regarding the extent of Amazon's past data scraping activities.
While Twitch has stated in its documentation that audio data may be utilized to refine speech-to-text models for automatic subtitles, the lawsuit contends that creators should not have their work commercially leveraged for AI training without explicit consent. Amazon, which acquired Twitch in 2014, has not provided a comment on the ongoing litigation.
Source: BBC News
暂无评论。快来发表第一条评论吧。