Urgent.News

What's breaking now, across thousands of outlets.

AI

What AI leaves out of the answer still counts

There’s a contradiction emerging in how publishers get paid when AI uses their content . The licensing deals that OpenAI, Meta, Amazon, and others have been signing these past few years appear to be slowing down, at least for the publishers still waiting on one. Publishers are speaking about this openly. Digiday hosted a London town hall recently, and one publishing executive said discussions…

What AI leaves out of the answer still counts

The conversation around publishers' payments for AI-generated content is shifting, with a new model gaining traction. While licensing agreements between major companies like OpenAI, Meta, and Amazon have slowed down, the demand for content continues to rise. Publishers are increasingly blocking AI bots, fearing the cost of handling countless requests.

This has led to the emergence of metered pricing, where payment is based on usage rather than a flat fee. Companies like Cloudflare and Ceramic.ai are implementing pay-per-use models, alerting publishers when a buyer wants their content, and allowing them to accept, decline, or negotiate. However, the effectiveness of this model hinges on determining how to measure usage.

While it's true that only a fraction of data used by AI systems contributes to the final answer, uncited sources still play a crucial role in shaping responses. Additionally, the rise of AI data brokers has introduced a middleman in the payment process. These companies, such as Exa, Parallel, and Tavily, crawl the internet and sell access to content, charging per request.

Parallel, in particular, has partnered with several publishers on payment through an index fee program, rewarding unique, hard-to-replace content. This new approach could potentially revolutionize the way publishers are compensated for their work in AI systems, but its success will depend on scaling and addressing the concerns of all parties involved.

Written by urgent.news from Fast Company's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at fastcompany.com →

More in AI

More from Friday 23 October →