How judges, not politicians, could dictate the rules of American AI

Its approach differs from that of other Western countries. While the EU is trying to proactively prevent the worst harms of AI, the US approach is more reactive. The US is waiting for harm to emerge before regulating, says Amir Ghavi, a partner at law firm Fried Frank. Gavi is representing Stability AI, the company behind the open-source image-generating AI Stable Diffusion, in three copyright lawsuits.

« This is a pro-capitalist position, » says Ghavi. “It promotes innovation. It gives creators and inventors the freedom to be a little bolder in imagining new solutions. »

Copyright and privacy class-action lawsuits could shed more light on how « black box » AI algorithms work and create new ways for artists and authors to be compensated for using their work in AI models, says Joseph Saveri , the founder of an antitrust and class action law firm, and Matthew Butterick, an attorney.

They are leading the lawsuits against GitHub and Microsoft, OpenAI, Stability AI and Meta. Saveri and Butterick represent Silverman, part of a group of authors who say tech companies have trained their language models on their copyrighted books. Generative AI models are trained using vast datasets of images and text pulled from the internet. This inevitably includes copyrighted data. Authors, artists and programmers say tech companies that erased their intellectual property without consent or attribution should compensate them.

“There’s a vacuum where the rule of law doesn’t exist yet, and we’re getting the law where it needs to go,” Butterick says. While the AI ​​technologies involved in the suits may be new, the legal issues surrounding them are not, and the team is relying on « good old copyright law, » he adds.

Butterick and Saveri point to Napster, the peer-to-peer music sharing system, as an example. The company was sued by record companies for copyright infringement and resulted in a landmark case over the fair use of music.

Napster’s deal paved the way for companies like Apple, Spotify and others to start creating new licensing-based deals, Butterick says. The pair also hope their lawsuits pave the way for a licensing solution where artists, writers and other copyright holders could also receive royalties for using their content in an AI model, similar to the system in place. in music. song sampling industry. Companies should also seek explicit permission to use copyrighted content in training sets.

Tech companies have treated copyrighted data publicly available on the Internet as subject to « fair use » under US copyright law, which would allow them to use it without first asking permission. Copyright holders disagree. Class actions will likely determine who is right, Ghavi says.