Authors have been fighting diligently to limit the encroachment of generative AI in courts and in contracts and now there is a new front in those battles as Amazon has quietly helped themselves to thousands of texts, with more planned soon. In October the company suggested, towards the bottom of a release focused on the refresh of their Kindle Scribe devices, they they would be “adding new AI-powered reading features that preserve the magic of reading on Kindle,” with “Story So Far” (a summary of what you had read so far) and “Ask this Book,” providing “spoiler-free answers to questions […]
AI
Bloomsbury Contracts With Google to Use AI Services
Bloomsbury has entered an agreement with Google Cloud to use the tech company’s AI services, including NotebookLM, Vertex AI and Gemini Enterprise. According to a release, “With an advanced AI infrastructure, Bloomsbury will benefit from data-driven and semantic search insights to improve trend analysis and drive book sales across Bloomsbury’s entire catalogue, while custom print prediction models can improve inventory management.” “Bloomsbury is pleased to be working with Google in this collaboration to demonstrate how cutting-edge technology can increase the discovery and sales of books, as well as transform engagement with content to improve learning outcomes,” ceo Nigel Newton said. […]
OpenAI Must Share Communications, Testify on Pirated Datasets
OpenAI has been ordered to provide information about their deletion of pirated ebooks that may determine the outcome of one of the copyright infringement lawsuits against the company. In the class action lawsuit brought by authors against OpenAI in New York’s Southern District, Judge Ona Wang has directed the tech company to disclose all of their in-house communications regarding why they deleted the Books1 and Books2 datasets that they used to train ChatGPT. Previously, OpenAI said that they deleted the datasets, obtained from LibGen, due to “non-use.” The company then backtracked, saying that all reasons for deletion were protected under […]
Huckabee’s AI Copyright Suit Against Bloomberg Moves Forward
In the latest class action lawsuit over copyright infringement in AI training, a group of authors including Mike Huckabee, US Ambassador to Israel and former Arkansas governor, is suing Bloomberg in a class action suit. The plaintiffs, David Kinnaman, non-profit The Relevate Group (which holds copyright for one of Kinnaman’s books), Tsh Oxenreider, Lysa TerKeurst, and John Blase, as well as Huckabee, are authors of mainly Christian nonfiction books. They allege that the media company violated copyright when it used the Books3 database to train its finance LLM, BloombergGPT. “Using data from Books3 enabled Bloomberg to create its LLM faster […]
OverDrive Sues OpenAI For Trademark Infringement Over Sora
OverDrive filed suit against OpenAI in a Northern Ohio Federal Court on Wednesday, alleging infringement of their trademark for the Sora brand, along with charges of unfair competition and deceptive trade practices. OverDrive’s Sora app for schools has been in use for over seven years, and was trademarked in 2022, while OpenAI’s Sora is their text-to-image creator first launched in December 2024. OverDrive writes in their complaint: “OpenAI’s recently launched AI text-to-video generation software app under the exact same name, Sora, using a confusingly similar icon, color palette, and visual identity, infringes OverDrive’s trademark rights and damages the integrity of […]
OpenAI Loses Copyright Suit In Germany, Whines In New York
For once justice is moving faster in Europe than here, as a German court became the first to find OpenAI guilty of copyright infringement. The case, brought by GEMA–the German music rights organization–accused ChatGPT of violating German copyright laws in reproducing lyrics by musician Herbert Groenemeyer and others. OpenAI’s defense was to argue that its own users were the ones infringing copyright when they prompted ChatGPT to generate the lyrics. Judge Elke Schwager found that OpenAI trained its LLM on copyrighted content from nine German songs. OpenAI will have to pay damages, though the amount was not specified. The ruling […]