According to newly revealed court filings, internal communications indicate that Meta's CEO, Mark Zuckerberg, authorized the use of the pirated dataset LibGen for AI training.
Authors like Ta-Nehisi Coates and Sarah Silverman argue that Meta misused their copyrighted works for its language model, Llama, claiming serious copyright infringements.
Court filings highlight that despite being flagged as 'pirated' by internal teams, the decision to use LibGen for training AI systems received executive approval.
Meta's authors are seeking to amend their complaint based on new evidence, introducing additional claims of computer fraud alongside copyright infringement.
Collection
[
|
...
]