From BigGo Finance · BigGo Finance · · 1m
US Court Orders Meta to Identify Data Decisions in AI Copyright Case
US court forces Meta to name all who approved pirated book data for AI training.
In brief
US court forces Meta to name all who approved pirated book data for AI training. US court orders Meta to identify individuals who approved obtaining pirated books from shadow libraries for AI training. Originally reported by BigGo Finance.
Court order
US Magistrate Judge Thomas Hixson ordered Meta to identify all who approved obtaining pirated books from shadow libraries and BitTorrent.
Data volumes
Plaintiffs allege Meta obtained 80.6 terabytes from LibGen and 35.7 terabytes from Anna's Archive.
Meta's defense
Meta claimed its early AI data buildout was informal with no authorized decision-makers.
Court rejection
Hixson rejected Meta's argument that decisions were made only technically by employees.
Decision-making requirement
Even team decisions require naming all individuals involved in approving data sources.
Internal messages
Plaintiffs cite internal messages showing decisions escalated to Mark Zuckerberg.
Unused data
Court seeks explanation for handling works downloaded but never used in model training.
At a glance
US court orders Meta to identify individuals who approved obtaining pirated books from shadow libraries for AI training.
