The firm wants to represent “real people whose information was stolen and commercially misappropriated to create this very powerful technology,” said Ryan Clarkson, the firm’s managing partner.
The lawsuit goes to the heart of a major unresolved question hanging over the surge in “generative” AI tools such as chatbots and image generators. The technology works by ingesting billions of words from the open internet and learning to build inferences between them. After consuming enough data, the resulting “large language models” can predict what to say in response to a prompt, giving them the ability to write poetry, have complex conversations and pass professional exams. But the humans who wrote those billions of words never signed off on having a company such as OpenAI use them for its own profit.
“All of that information is being taken at scale when it was never intended to be utilized by a large language model,” Clarkson said. He said he hopes to get a court to institute some guardrails on how AI algorithms are trained and how people are compensated when their data is used.
The legality of using data pulled from the public internet to train tools that could prove highly lucrative to their developers is still unclear. Some AI developers have argued that the use of data from the internet should be considered “fair use,” a concept in copyright law that creates an exception if the material is changed in a “transformative” way.
The question of fair use is “an open issue that we will be seeing play out in the courts in the months and years to come,” said Katherine Gardner, an intellectual-property lawyer at Gunderson Dettmer, a firm that mostly represents tech start-ups. Artists and other creative professionals who can show their copyrighted work was used to train the AI models could have an argument against the companies using it, but it’s less likely that people who simply posted or commented on a website would be able to win damages, she said.
The suit also adds to the growing list of legal challenges to the companies building and hoping to profit from AI tech.
A class-action lawsuit was filed in November against OpenAI and Microsoft for how the companies used computer code in the Microsoft-owned online coding platform GitHub to train AI tools. In February, Getty Images sued Stability AI, a smaller AI start-up, alleging it illegally used its photos to train its image-generating bot. And this month OpenAI was sued for defamation by a radio host in Georgia who said ChatGPT produced text that wrongfully accused him of fraud.
OpenAI isn’t the only company using troves of data scraped from the open internet to train their AI models. Google, Facebook, Microsoft and a growing number of other companies are all doing the same thing. But Clarkson decided to go after OpenAI because of its role in spurring its bigger rivals to push out their own AI when it captured the public’s imagination with ChatGPT last year, Clarkson said.
