Updated July 15, 2023:
Google has to go to court, too
Google, Deepmind, and Alphabet are in the dock alongside Meta and OpenAI: Google faces a total of ten allegations in a class action lawsuit filed in California federal court by eight anonymous individuals, including two minors. The main allegation is that Google stole massive amounts of personal information to train its Bard chatbot.
It has very recently come to light that Google has been secretly stealing everything ever created and shared on the internet by hundreds of millions of Americans. Google has taken all our personal and professional information, our creative and copywritten works, our photographs, and even our emails—virtually the entirety of our digital footprint—and is using it to build commercial Artificial Intelligence (“AI”) Products like “Bard ,” the Google chatbot recently released to compete with OpenAI’s “ChatGPT.” For years, Google harvested this data in secret, without notice or consent from anyone.
OpenAI is also under scrutiny by the FTC, primarily for ChatGPT hallucinating wrongs and harmful stories about individuals.
Updated July 10, 2023:
Authors on Meta and OpenAI for possible AI copyright infringement
Behind the lawsuits are Joseph Saveri and Matthew Butterick, the legal team already suing Github Copilot and Stable Diffusion for copyright infringement. The lawyers leave no doubt that they consider the unsolicited use of code and images from the Internet for AI training to be theft.
They are now applying this logic to authors whose texts were part of the training material: “Since the March 2023 release of OpenAI’s ChatGPT system, we have heard from writers, authors, and publishers concerned about the system’s uncanny ability to generate texts similar to those in copyrighted textual material, including thousands of books.”
Saveri and Butterick are representing authors Paul Tremblay and Mona Award in a lawsuit against OpenAI (see below), alleging that ChatGPT, GPT-3.5 and GPT-4 “remix” the works of thousands of book authors without consent, payment or reference.
The attorneys are also representing comedian Sarah Silverman, Chris Golden and Richard Kadrey in a second class action lawsuit against OpenAI over the same issue. The three authors are having Saveri and Butterick file another class action against Metawhose LLaMA language model was also trained with copyrighted books.
LLaMA has provided the technical basis for numerous open-source models, some of which are marketed commercially. Meta plans to make LLaMA v2 the center of open source development as well.
Original article, July 5, 2023:
OpenAI faces two US class action lawsuits for web scraping and copyright infringement
OpenAI is facing laws over web scraping and copyright infringement. One class action lawsuit accuses the AI company of misusing personal data on an unprecedented scale. Another lawsuit alleviates that ChatGPT was trained on copyrighted books without the authors’ consent.
OpenAI’s ChatGPT and DALL-E AI programs were trained on “stolen private information” from millions of Internet users, including children, without their consent, the laws filed in federal court in San Francisco say. OpenAI is also accused of secretly violating terms of service and federal and state privacy and property laws. Among the defendants is Microsoft, one of OpenAI’s major investors.