Automobile Photography Company Sues Universities for Using Images in AI Training Models
EVOX Productions, LLC ("EVOX") recently filed a Complaint against several universities, namely Stanford University, Rice university, University of Michigan, and the Baylor College of Medicine (collectively "Defendants"). According to the Complaint, in 2020 a group of faculty from each of the Universities worked together on a project and acted as a "Consortium" to appropriate nearly 8,000 images from EVOX's catalog for use in AI training data. After the model was trained, the Consortium allegedly hosted the images on the Stanford website as free downloads for anyone to use as AI training data.
In 2023, the same group of researchers supposedly posted a second set of AI training data to the Stanford website, this one a smaller set of 225 images. Upon discovering this set, EVOX and the Defendants entered into a Tolling Agreement. A tolling agreement is where parties agree to suspend the statute of limitations for a given disagreement so that each party can have extended time to research and decide whether or not to continue with litigation. Thus, the statute of limitations has not expired for the possible 2020 infringement and those claims are available for EVOX to pursue alongside the possible 2023 infringement.
AI training data has been a hot topic in recent years, with almost no solid litigation and precedent to help copyright holders and AI researchers understand what the limitations of copyright and fair use are in the field. As some of the Defendants are private universities and cannot seek Sovereign Immunity, we will likely see them try to apply Fair Use, which could provide much needed precedent and instruct the future of AI in the United States. We will provide updates in this exciting case as it moves forward in the litigation process.