Big trouble for Meta, the brains behind Facebook and Instagram.
They’re in hot water as authors like Sarah Silverman and Michael Chabon team up in a lawsuit, claiming Meta used copyrighted books without permission to train its AI, called Llama.
Even with Meta’s legal team warning about the risks, the company allegedly went ahead. Now, chat logs show Meta researcher Tim Dettmers talking about getting the dataset, with the legal team expressing concerns.
This legal drama started in the summer, and recently, a judge dismissed part of Silverman’s lawsuit. Now, the authors are tweaking their claims, making things more interesting.
It’s not just a Meta problem, if these lawsuits succeed, it could shake up the AI world. Companies might face more scrutiny and costs. Also, new rules in Europe might make AI companies spill the beans on their training data, adding more legal risks.
At the center of it all are Meta’s Llama models, especially the latest one, Llama 2. While the first version spilled the beans on its training data, Llama 2’s details are still a mystery.
What are your thoughts? Let us know!
