The Real AI Battleground Is 200 Years Old

For most of the AI boom, attention has centered on the companies building models, chips, and applications. OpenAI, Anthropic, Google, and NVIDIA have dominated the conversation. Academic publishers, for one, have not.

The shift is visible in the recent moves of Wiley, the 219-year-old scientific publishing company. Over the past year, Wiley partnered with OpenEvidence, Microsoft, Anthropic, Amazon Web Services (AWS), and IQVIA. It acquired Emerald Publishing for $452 million and nearly doubled its AI-related revenue in two years.

Other publishers are moving in the same direction. Companies that spent decades collecting and validating knowledge are finding that those archives have a new kind of buyer, as enterprises search for trusted data to train and power AI systems.

Publishers Spent Decades Building What AI Suddenly Needs

Generative AI has made data quality a strategic issue. Large language models can generate convincing answers on almost any subject, but enterprise and scientific applications require a higher standard. Pharmaceutical companies developing drugs and engineers designing products cannot rely on information scraped from the internet.