Publishers Launch Meridian for Licensed AI Access

Three university presses have founded Meridian, a licensing service run by Cashmere that gives students and researchers authorized AI access to scholarly content.

Updated on Nov. 3, 2026 in Artificial Intelligence

Isometric editorial illustration of stacked bound journal volumes beside a brass key on a stone plinth, with a beam of light leading into an arched doorway.
Three university presses launched Meridian, a licensing service run by Cashmere that gives students and researchers authorized AI access to scholarly content, with a limited test run beginning this fall. AI Illustration. Upload story photo >

Live Poll

Should AI companies have to pay publishers when their tools use books and articles?

Cambridge University Press, Johns Hopkins University Press, and Princeton University Press founded Meridian, an AI licensing service run by Cashmere. Its limited test run with libraries and institutions begins later this fall.

Why it matters

Publishers and libraries are pursuing shared rules for AI access to scholarship before outside technology platforms set those rules. Meridian lets publishers connect to AI tools through one arrangement rather than a separate deal with each AI company.

Meridian uses the Model Context Protocol, an open standard developed by Anthropic, to connect AI tools to peer-reviewed content and cite sources. Publishers build a single connection to Meridian instead of separate connections for each AI company.

The players

Cambridge University Press

A major academic publisher whose founding participation in Meridian signals a coordinated effort among scholarly presses to set terms for AI access to their content.

Johns Hopkins University Press

The university press of Johns Hopkins University, which joined the founding group that created Meridian as a publisher-controlled alternative to unregulated AI scraping of scholarship.

Princeton University Press

The scholarly press of Princeton University, which co-founded Meridian alongside Cambridge and Johns Hopkins University Presses.

Cashmere

A Salt Lake City-based company that built the licensing technology behind Wiley's deal with Perplexity and now runs Meridian.

The details

Meridian gives students and researchers authorized access to scholarly books and journals that libraries already license, and content accessed through it cannot be used to train AI models or resold to other companies. When a researcher asks an AI tool a question, Meridian lets the tool pull its answer from peer-reviewed content and sends the reader through their library's subscription to the final published version.

Timeline

  1. July: Queen and Woahn called for shared publisher-controlled AI rules.

  2. Earlier this year: Cashmere raised $5 million from investors led by Reach Capital.

  3. This month: Founding publishers will load their catalogs onto Meridian.

  4. Later this fall: A limited test run with libraries and institutions begins.

The Tech Race

Meridian's use of the Model Context Protocol places academic publishers on the same open-standard footing that AI developers have been building around. It arrives as publishers seek terms for AI access before outside platforms define them.

Students and researchers at participating libraries may gain authorized AI-assisted access to scholarly books and journals they already have licensed. Meridian bars AI training on that content and reselling it, which limits how the material can be reused.

The takeaway

Libraries and researchers should watch how Meridian's limited test run handles citation and licensing before assuming AI tools will respect publisher terms. Authorized access will depend on whether a participating library subscribes to the content.

What happens next

Founding publishers will load their catalogs this month, and a limited test run with libraries and institutions begins later this fall. The council will become Meridian's board once it incorporates as an independent nonprofit.

Further reading

Explore more on Artificial Intelligence developments shaping how scholarly content is licensed and accessed.

Live Poll

Should AI companies have to pay publishers when their tools use books and articles?