In the race to advance artificial intelligence (AI) capabilities, tech giants are fervently seeking new data sources. Meta, in particular, has grappled with this challenge, with executives convening almost daily last year to address the issue, as reported by The New York Times.
As AI technologies grow more sophisticated, companies are intensifying their efforts to acquire data, potentially raising concerns about copyright infringement. OpenAI, for instance, faced allegations of utilizing YouTube content to train its video generator, Sora, although the company’s CTO, Mira Murati, refuted these claims.
According to The Times, Meta’s meetings involved discussions about various strategies, including the prospect of acquiring publishing house Simon & Schuster, purchased by private equity firm KKR for $1.62 billion in August. Alternatives such as paying $10 per book for full licensing rights to new titles were also proposed.
Prior to these meetings, Meta had already compiled summaries of numerous books, essays, and online works, some of which contained copyrighted material. During discussions, concerns were raised about the ethical implications of collecting intellectual property without proper licensing agreements, prompting silence from attendees, as per The Times.
Meta has yet to respond to Business Insider’s request for comment.
Ultimately, executives drew on the precedent established in Authors Guild vs. Google, a landmark 2015 case. In this case, the Supreme Court declined to intervene, affirming a lower court ruling that allowed Google to scan and digitize books for Google Books under fair use guidelines. Meta’s legal team asserted that the company could train its AI systems within the same framework, The Times reported.