Has your paper been used to train an AI model? Almost certainly

Academic publishers are increasingly selling access to research papers to technology firms for training artificial intelligence (AI) models, raising concerns about copyright and the lack of consultation with authors. Notable deals include Taylor & Francis's $10 million agreement with Microsoft and Wiley's $23 million deal with an unnamed company. Experts suggest that if a research paper hasn't been used yet, it likely will be soon, as AI models require vast amounts of data, often sourced from online content. The trend is growing, with other media outlets also offering their content for AI training. While some AI developers maintain open data sets, many keep their training data secret, complicating the ability to determine if specific papers were used. Legal disputes are emerging, such as The New York Times suing Microsoft and OpenAI for unauthorized use of its content. Although some researchers are indifferent to their work being included in AI training, many express frustration over the lack of control and fair compensation for their contributions. The ongoing discussions highlight the need for a balanced approach to the use of academic work in AI development, ensuring that authors' rights are respected while fostering innovation in AI technologies.

- Academic publishers are selling research papers to AI firms, raising copyright concerns.

- Major deals include Taylor & Francis with Microsoft and Wiley with an unnamed company.

- The trend of using academic papers for AI training is expected to grow.

- Legal disputes are emerging over unauthorized use of copyrighted content for AI training.

- Researchers express frustration over the lack of control and compensation for their work.