Sarah Silverman Joins The Quest Against AI

Tech

Jul 11

Sarah Silverman has joined novelists Christopher Golden and Richard Kadrey to file suit against OpenAI and Meta for copyright infringement. The trio cited that their published works were obtained without their permission. OpenAI and Meta claim that they use “Books1” to train ChatGPT. The “Books1” dataset is approximately the same size as Project Gutenberg. Project Gutenberg, of course, is a large book repository where consumers can choose to read from a wide selection of work for free online. Attorneys for Silverman, Golden, and Kadrey argue that the “Books2” dataset used in AI training is wherein lies the copyright violation as it is much too large for the companies to obtain anywhere else but the “shadow libraries.”

Shadow libraries often house illegally obtained copyrighted work, examples of such being Library Genesis and Sci-Hub. People can pirate this material under the radar, with some libraries even offering bulk package deals to larger-than-average orders. Silverman’s legal team doubled down on their evidence when the team asked the chatbot about Silverman’s 2010 memoir titled, The Bedwetter. Not only was the AI able to summarize the book, but it recited some pieces verbatim. This is far from the first time that Open AI has been accused of violating copyright as well as both federal and state privacy laws in the name of training artificial intelligence. Moreover, in June, the company was served with two additional complaints, one of which was a class action suit against the company. OpenAI is responsible for both ChatGPT and DALL-E.

READ NEXT:

Featured