TapTechNews August 14th news, the presiding judge of the local court in the US ruled yesterday (August 13th) that the collective lawsuit case of AI image training can continue to be advanced, but some claims were rejected.
TapTechNews learned from the report that the plaintiffs are composed of many artists, and the Defendants are StabilityAI, Midjourney and other AI-related companies. The Plaintiffs accused them of illegally using copyrighted works for training AI.
The Plaintiffs of this collective lawsuit said that currently, the data sets used by many mainstream AI text-to-image services contain the copyrighted works they created.
Judge William Orrick approved the additional inducement copyright infringement claim against the Stability company and accepted the Plaintiffs' claims of copyright claims against DeviantArt (the company uses a model based on StableDiffusion), RunwayAI (the startup behind StableDiffusion), and copyright and trademark infringement claims against Midjourney.
However, the judge rejected the accusation that the AI generator deleting or changing copyright management information violates the Digital Millennium Copyright Act, and also rejected the accusation that DeviantArt violated the service terms by allowing users' works to be used in the AI training data set.
In the accusation against Midjourney company, 4700 artists participated in the collective lawsuit, believing that the company had misleading content and agreed to share content without their knowledge or consent.