Not a day without trial: OpenAI turned out to be a fan of science fiction and pirates
OpenAI, you steal my books – No, I process them.
Company Open AI faced another lawsuit. This time, science fiction writers Paul Tremblay and Mona Awad filed class action against OpenAI in federal court, accusing the company of copyright and privacy infringement.
Tremblay and Awad claim that OpenAI used their books in training ChatGPT without obtaining proper consent, acknowledgment of authorship or compensation. They believe that ChatGPT’s ability to accurately abstract their works is evidence that their books have been copied and included in the OpenAI language model without permission.
The lawsuit points to a 2020 OpenAI article that talks about using “two sets of books from the internet” to train ChatGPT. The lawsuit authors suggest that one of these datasets, which includes more than 290,000 titles, could come from shady online libraries that publish copyrighted works.
According to the lawsuit, OpenAI integrated Spotify, snapchat, Stripes, Slack, and Microsoft Teams to their systems to collect personal data from users, including images, locations, music tastes, financial data, and private messages. Collecting such data violates the terms of service of those platforms.
In a separate lawsuit, OpenAI is also accused of illegally collecting personal information from the Internet using its ChatGPT and DALL-E AI models. The lawsuit alleges that such data collection violates a number of state and federal privacy laws.
The lawsuits filed could have far-reaching implications for OpenAI and the entire AI industry. If the court sides with the plaintiffs, OpenAI could face heavy financial penalties that could affect its financial stability and fundraising. In addition, lawsuits can damage a company’s reputation, leading to increased regulatory scrutiny and a need for greater transparency.
The impact will not only affect OpenAI – companies using ChatGPT or other OpenAI products may consider partnering with OpenAI to protect their reputation and protect user privacy. Moreover, if courts determine that using copyrighted material to train AI models is infringing, OpenAI and other companies may be forced to reconsider their data collection practices.
Recall that recently Microsoft and OpenAI were at the center of a legal scandal because of their artificial intelligence products, which, according to 16 anonymous plaintiffs, illegally used and distributed their personal data without proper notice or consent.
In addition, the Norwegian Consumer Protection Council (Forbrukerradet) warned about the risks of generative AI and urged politicians and regulators not to trust the tech giants to ensure that the technology is secure and reliable.