Meta May Have Used Unpublished Books to Train AI Models

In recent days, the controversy over whether Meta might have used unpublished books to train its AI models has resurfaced. According to reports, The Atlantic released a new tool last week designed to search the LibGen database, and Meta has been accused of using data from this database to train its AI models.

During an investigation into the matter, an author discovered that their upcoming work had already appeared in the LibGen database. This collection was scheduled for release on July 1st, yet it seems Meta had already accessed its contents and utilized them for AI model training.

It's worth noting that digital copies of unpublished works are typically only available through legitimate channels like NetGalley and Edelweiss. These platforms impose strict restrictions and guidelines on user behavior, clearly stating that users have very limited permissions regarding the use of unpublished works.

This incident has once again drawn public attention to issues of data privacy and intellectual property protection. Unpublished works, as the intellectual output of authors, deserve full respect and protection. As a tech giant, Meta should set an example in complying with laws and respecting intellectual property rights.

Currently, whether Meta indeed used unpublished books to train its AI models, and whether such actions are legal and compliant, still require further investigation and scrutiny. Nevertheless, this event serves as another reminder of the critical importance of data privacy and intellectual property protection.