OpenAI, Microsoft, and Google are brokering data deals with news outlets. Are music companies next?
Reports from major news outlets like the Financial Times suggests these companies are meeting with news executives about the potential copyright issues presented in generating AI content. Large language models like ChatGPT and generative AI models like Midjourney and Stable Diffusion are trained on existing content. Stable Diffusion is facing a lawsuit from Getty Images due to the Getty watermark showing up in AI-generated images—which means Getty images were part of the dataset to train Stable Diffusion.
The FT reports that executives from News Corp, The New York Times, The Guardian, and Axel Springer have all discussed a subscription-style fee for content used to develop artificial intelligence models. A deal struck with these news organizations may set up the blueprint for similar copyright issues with generative AI. So what does that pricing model look like? It’s too early to tell, but early reports suggest anywhere from $5 million to $20 million a year to access content.
Talks include discussing creating a ‘quantitative model‘ for access to data—similar to what is already employed by the music industry. Radio stations, night clubs, and streaming services all pay record labels when a track is played, which is what news outlets want to see from AI companies. Each time a piece of information is accessed—that owner gets paid. But in order to set up a model like this, it would require said companies to disclose the usage of media content in their current operations—which these companies are not willing to do.
OpenAI CEO Sam Altman even suggested that his company would cease operating in the EU if the nation bloc legislated regulations on its operations. The EU AI Act would force companies like OpenAI to disclose copyrighted material used to train its systems.
“If OpenAI can’t comply with basic data governance, transparency, safety, and security requirements, then their systems aren’t fit for the European market,” says Dutch MEP Kim van Sparrentak, who has helped draft the current EU AI Act as it stands.