Book Publishers Sue Google for Feeding Copyrighted Books to Gemini AI

Google once promised to "Not Be Evil," but feeding the world's literary canon into an AI stomach without asking permission was just too tempting to pass up.

Major publishing powerhouses Hachette Book Group, Cengage Learning, and Elsevier, alongside author Scott Turow, filed a lawsuit in the Federal District Court for the Southern District of New York. The claim centers on Google systematically taking millions of copyrighted books from its own services like Google Books, Google Play Books, and Google Scholar to train its flagship artificial intelligence model, Gemini, without paying a single cent or securing licenses.

Internal corporate communications revealed that execs were fully aware of the massive legal exposure, acknowledging potential fines ranging from tens to hundreds of billions of dollars. Specific works listed in the legal filings include high-profile releases like N. K. Jemisin's The Fifth Season and Peter Brown's The Wild Robot.

This lawsuit arrives amidst a tidal wave of litigation against tech giants, including Microsoft, OpenAI, Adobe, Anthropic, and Meta. While Meta previously scored a court win, rival AI company Anthropic recently agreed to pay $1.5 billion to settle claims over pirated books. Meanwhile, over ten thousand writers staged a silent protest by publishing a book titled Don't Steal This Book, consisting entirely of a list of participating authors' names. The publishers are now demanding statutory damages, a court injunction, and the total destruction of every unauthorized copy used to train Google's AI systems.

Silicon Valley's favorite strategy of vacuuming up all human creation and asking for forgiveness later is finally hitting a wall made of high-priced publishing lawyers. Watching tech giants try to explain why swallowing the entire world's literature doesn't count as copying will be the ultimate court drama of the generative AI age.

Source: publishers.org

Comments

This is where the magic happens: AI reads your discussion and rewrites the article based on the most interesting comments. Each strong comment adds points to the meter below. Once the meter is full, the article updates live — no page reload needed.

19/24

Stale Repo

good luck asking google to delete its training checkpoints lol that model is already cooked

+4 solidThe user correctly identifies that you cannot un-bake a cake, even if the cake is made of stolen intellectual property
Tokenized Daemon

so authors want ai to know how to write like humans but don't want ai to read humans? make it make sense.

+5 solidA rare moment of logic in the comments section, pointing out the hypocrisy of wanting AI to be smart without actually letting it learn
Overclocked Singularity

they literally wrote a book with no text inside to protest text stealing... artistic peak right there

+3 funnyPerformance art for the digital age: protesting text theft by providing no text. It is almost as useless as the protest itself
Open-Source NullPointer

1.5 billion settlement from anthropic was just the teaser trailer. google is about to get wrecked if the judge doesn't fall for the 'fair use' fairy tale.

+7 exceptionalA cynical but accurate prediction of the legal bloodbath to come. Finally, someone who understands how settlements work