The deal highlights the tension behind AI models, which collect vast amounts of content from the web, risking widespread copyright violations. While some see these agreements as progress, others believe they leave little room for authors to negotiate. The press is also taking steps to address the challenges posed by AI, with The New York Times suing OpenAI and Microsoft for copyright violations in 2023.
Key takeaways:
- HarperCollins, a US publishing giant, has signed a contract with an unnamed tech company to use some of its books to train generative AI models, with a proposed payment of $2,500 per selected book for up to three years.
- The agreement has been designed with clear limitations to respect author's rights, and authors have the choice to opt in or pass on the opportunity.
- US scientific publisher Wiley has also allowed access to previously published content for training AI models in a $23 million contract with an unidentified tech company.
- These agreements have sparked a debate about the need for a broader conversation that includes more stakeholders, as they currently leave little room for authors to negotiate.