Microsoft pilots new content marketplace for AI training: What it means for publishers

Microsoft’s PCM platform comes amid a long-running stand-off between publishers and big tech companies which has only intensified with the rise of AI. The current AI boom has been largely fueled by large language models (LLMs) that have been trained and developed by ingesting vast amounts of training data scraped from all corners of the internet, including publishers’ websites, without authorisation.

In response, publishers such as The New York Times have filed copyright infringement lawsuits against tech companies such as Microsoft and OpenAI. In India, publishers — members of the Digital News Publishers Association (DNPA), including The Indian Express, among others — have mounted a legal challenge against OpenAI over the “unlawful utilisation of copyrighted material”.

On the other hand, several major publishers have inked lucrative deals with AI companies to licence their content for AI training purposes.

“The open web was built on an implicit value exchange where publishers made content accessible, and distribution channels – like search – helped people find it. That model does not translate cleanly to an AI-first world, where answers are increasingly delivered in a conversation,” Microsoft said.

“At the same time, much of the authoritative content lives behind paywalls or within specialised archives. As the AI web grows, publishers need sustainable, transparent ways to govern how their premium content is used and to license it when it makes the most sense,” it added.

The Windows maker said its newly unveiled PCM platform has been developed in partnership with US-based publishers such as Vox Media, The Associated Press, Condé Nast, People, and others. To show that training models on premium content improves the quality of their responses, Microsoft said it grounded specific responses of its Copilot AI chatbot with licensed content and ran experiments to validate assumptions before scaling. Its testing revealed that premium content meaningfully improved the quality of Copilot’s responses.

The company further said it is looking to onboard Yahoo and other partners as it continues to pilot test the PCM platform.

India’s proposed licensing regime

Last year, a government-constituted committee led by the Department for Promotion of Industry and Internal Trade (DPIIT) proposed a new framework which would require all AI companies to pay royalties to creators for using copyrighted work under a mandatory blanket licence.

The flat royalty rates would apply retrospectively and would be prescribed by a government-appointed committee, according to the working paper drafted by the DPIIT-led committee. The collection and distribution of royalties to creators would be overseen by a new umbrella industry body called the Copyright Royalties Collective for AI Training (CRCAT).

Notably, the committee rejected direct, voluntary content licensing models between AI developers and individual companies, saying it would lead to high transaction costs, long negotiations, and unequal bargaining power that disadvantage small creators and startups, without offering broad, dependable access to training data.