Business & Enterprise2 min reading time

Meta tries to scapegoat staff who downloaded torrents to spare execs

Ars Technica
Read the full article
Meta denied that executives formally approved the use of pirated datasets for training AI models like Llama 1 and 2, attributing decisions to informal or collaborative staff actions. A court urged Meta to provide detailed information on who authorized or contributed to downloading such data and what happened to unused pirated works.

More in Business & Enterprise

AI delivers ‘zero financial payoff’ to most companies, Apollo economist says

Covered by 2 sources

OpenAI cuts its valuation target to $1.25tn as it nears a $30bn round

The Next Web