Going further, Meta accused Entrepreneur Media of “speculation,” stating that there is “no basis” to suspect that “someone at Meta must have ‘approved’ a ‘decision’ to conduct a seeding activity for which specific mitigations were put in place to prevent.”
Instead of naming staff and providing details such as their “role, approximate dates of involvement” and “whether the person made or contributed to the decision and/or authorized, ratified, directed, permitted or approved it,” Meta suggested only providing names. (Hixon disagreed that that would be appropriate.)
To avoid naming executives, Meta explained that it wasn’t common practice for leaders to make “formal” decisions when acquiring data sets for early AI training on models like Llama 1 and 2. Only later, Meta said, did that “initial informal approach” give way to more structured review and approval processes.
Disbelieving Meta’s arguments, Hixon said the authors did not limit their request to formal decisions, while agreeing that Meta should do more to answer questions about who “made, authorized or approved any decisions to obtain or download the shadow libraries.”
Another tactic Meta used was to claim that torrent approvals came from “collaborative” conversations, so there was “no single decision maker.” Again, Hixon suggested that Meta was reading the authors’ request too narrowly. His motion “did not assume that Meta had a single decision-maker; it asked for all decision-makers,” Hixon said.
Finally, Meta must also answer questions about what happened to pirated works that were not used for AI training, Hixon said.
Lawyer for Meta and Entrepreneur Media did not immediately respond to Ars’ request for comment.
