Silicon Valley’s unlikely ally to defend generative AI

Anti-copyright activists fought the film and music industries. Now they’re fighting artists who want to stop generative AI.

Silicon Valley’s unlikely ally to defend generative AI
Photo: Unsplash/Bermix Studio

When ChatGPT was released in November 2022, many artists and publishers quickly realized another technological threat was on the horizon. Boosters were enthusiastic about how generative AI tools could churn out writing and images that would, in the process, devalue the work of writers, journalists, illustrators, and many other professions. But it all depended on the ability of those companies to take the collective knowledge of billions of people from the open web to feed into their artificial intelligence (AI) models.

Immediately, many of those professions who felt threatened saw an opportunity they could seize to present a legal speed bump for the seeming inevitability of AI adoption. Over the past year and a half, a series of groups that include visual artists, authors, and news organizations have launched copyright lawsuits directed at that vulnerability. The US Copyright Office has also refused to grant copyright protection to works created with generative AI tools — a campaign that has been supported by groups like the Writers Guild of America, SAG-AFTRA, and the Authors Guild.

If governments or the courts affirm the demands of artists and media groups, AI companies would have to remove copyrighted works from their training data or develop complex new systems to identify and compensate the owners of those copyrights. That’s something they want to avoid at all costs. In recent months, OpenAI has been proactively making deals with news publishers to gain access to their archives, but the industry isn’t stopping there.

Last week, tech lobby group Chamber of Progress launched a campaign called “Generate and Create” to ensure it remains legal for AI companies to train their models on copyrighted works. The group represents Amazon, Apple, Meta, Google, and many others. Their argument is not only that generative AI “lowers barriers for producing art,” but that training on existing works should be considered fair use. They’ve found an unlikely ally in the argument: anti-copyright activists. After taking on film and music industry lobbyists in service of the little guy in the 2000s, they’ve flipped as their libertarian politics push them defend some of the biggest and most exploitative companies in the world.