Nearly 300 authors filed copyright infringement claims against Meta Platforms on October 19 in the U.S. District Court for the Northern District of California, alleging the company trained its AI models on pirated copies of their books and, more recently, on physical books it bought and scanned without a license. Rather than join a class action suit, in which a few named plaintiffs sue on behalf of everyone, each author is pursuing an individual claim.
Source: Hundreds Sue Meta for Copyright Infringement, Rejecting Class Action Payouts
The model we’ve quietly relied on for 20 years, outsourcing product development to whoever can afford to be sued, is no longer fit for purpose. So, the license can’t sit with the creator. It has to sit at the engine layer, carried by the platform as infrastructure, the way blanket rights sit under radio. Which means the deals we should be doing right now are with the engine makers, before the defaults harden.
Spotify‘s quest to curb the chaos AI-generated music has brought onto the platform has prompted it to remove 50 million AI slop tracks from users’ recommendations this year, a new figure shared Tuesday as part of the company’s public insights into how it’s navigating the AI era. The 50 million tracks the streamer removed from recommendations were in addition to more than 75 million tracks Spotify pulled completely after being flagged as spam.
The Writers Guild of America East has struck a partnership with the AI shielding tool Cloak, a service that aims to help writers safeguard their work from training AI models. The effort responds to years of training from those models on screenplays without explicit permission, which has sparked questions over the legality of using those scripts without permission. The union has maintained that using scripts to train AI models does not constitute fair use.
The International Confederation of Music Publishers (ICMP) has accused AI companies of trying to “strong-arm” Australia’s government into loosening the country’s copyright rules on AI training. ICMP warned that, of the dozens of countries AI firms have lobbied for weaker copyright protections, Australia appears to be the only one considering the idea. ICMP Director General John Phelan said AI companies were asking policymakers to “legitimize larceny.”
News publishers have until 30 October to decide whether to leave Meta social platforms such as Facebook and Instagram or hand over over AI content access to the tech giant under new terms and conditions. The social media giant has given notice of updated terms and conditions which give it sweeping rights to scrape content from any publisher that uses Instagram and Facebook and reuse it.
The labels made their case in a filing in Boston federal court last Tuesday (October 13), opposing Lightspeed’s motion to quash their subpoena. Among the topics they want a Lightspeed representative to answer questions on: the “legal risk of training data,” as discussed in a document Lightspeed has already produced in the case. The same Lightspeed document discusses a “two sided music platform,” according to the subpoena.
Generative artificial intelligence differs fundamentally from conventional photographic manipulation. Photoshop alters an existing image. A generative model can create an image of an event that never occurred, involving people who were never present in a location that may itself never have existed. 
Meta will have to name the people who greenlit its torrenting of pirated books. A federal magistrate judge has ordered the company to identify everyone who approved sourcing books from LibGen and Anna’s Archive. This goes beyond the employees who ran the downloads and includes whoever decided whether to seed the books to others.
With an eye on challenging Spotify and Amazon’s Audible, YouTube is reportedly preparing to add audiobooks to its Premium subscription offering. That possible audiobooks expansion surfaced in a Wall Street Journal report, though at the time of this writing, Google/Alphabet hadn’t commented publicly on the rumored buildout. But per the outlet, the tech giant’s been engaged in talks with (book) publishers in recent months ahead of a possible launch sometime next year.