The filing corresponds to a lawsuit against Meta regarding AI copyrights. This one was launched by Kadrey who accused Meta Platforms of not respecting the copyright ownership of authors and various other intellectual property holders. So many AI firms claim that it’s fair to use such material but the plaintiffs disagree as it’s a clear violation of their copyrights.
The latest filings were rolled out to the court that entail partial transcripts of various Meta employees that prove how strongly many staff members felt about AI training data licensing for books not being a very viable option for the firm.
As per one transcript, Meta’s outreach to different publishers was met with a slower update in both engagement and interest. While they cannot recall the entire list, they speak about scouring various top publishers of the internet as per the transcript obtained. They also didn’t get many contacts or feedback from cold calls to try and establish the right kind of contacts.
While quite a few did engage, it was not the same case for all. The court transcripts show how Meta paused quite a few AI-based book licenses during the start of 2023 after coming across major logistical and timing setbacks. Many fiction book publishers didn’t have rights to content that Meta was entering into a licensing deal with.
It was soon pointed out that this was very true for the fiction category where many people representing themselves were doing it for rights they didn’t have in the first place, Meta’s spokesperson mentioned. So as per the company, there was no need to engage with such authors.
Meta has in the past also paused its licensing efforts linked to AI development as per this transcript. The company is well aware of licensing efforts that were paused once in the past before but that’s because they wished to design their own solution.
The complaint argued how Meta makes use of shadow libraries featuring pirated e-books for training different company AI models. This includes its Llama open model series. As per the complaint, Meta could have secured some of these libraries through torrenting. The latter has to do with distributing files through the web and needs torrents to seed or upload files they’re trying to obtain at the same time. However, the plaintiffs argued that it was a form of copyright infringement.
Image: DIW-Aigen
Read next: TikTok Returns To App Stores Of Apple and Google In America After President Trump Delays Ban
by Dr. Hura Anwar via Digital Information World
No comments:
Post a Comment