UK news publishers Guardian Media Group, Sky News, and dmg media announced today that they have all signed up to support ProRata.ai, a startup which seeks to ensure that publishers and creators are credited and compensated when their work is surfaced by generative AI tools, on a per-user basis. Dmg media specifically said it has invested in an equity stake in the platform, becoming the first UK news publisher to do so.
Reach PLC, Mediahuis, Mumsnet, Hello!, Prospect, and Exponential View have all also signed up as partners in the UK and Ireland, joining existing supporters including Axel Springer, Fortune, Time, The Atlantic, and Universal Music Group.
Signed up publishers will all see their content used to inform answers generated by ProRata’s ‘answer engine’, which is due to launch this year. Fifty percent of ProRata’s revenues, generated by a mix of advertising and subscription fees, will then be handed to partnered publishers, with the amount received by each publisher dependent on the extent to which their content has been used.
“ProRata continues to gain significant momentum,” said Bill Gross, CEO of ProRata. “Our AI technology is the only one that credits and compensates creators while providing consumers with highly accurate search results. We have had hundreds of content owners and media companies reach out to us from around the world who are interested in piloting our technology. Stealing and scraping content is not a sustainable path forward.”
Retracing AI’s steps
In some cases, attribution and compensation for publishers’ work in generative AI tools is theoretically fairly straightforward. If a user asks a question about a specific news story, an AI tool may in some cases surface direct quotes or details from several outlets, and can then provide a source for each one. Or where an AI chatbot could choose to base an entire answer off of one publisher’s work, making attribution even simpler. A lot of the licensing deals being agreed between the AI giants and news publishers work along these lines.
For many AI queries however, things aren’t so simple. For complex questions, a wide range of sources with overlapping content may be used, making the contribution of each individual publisher to the final answer unclear.
This is where ProRata says its technology fits in. The company has created technology which analyses AI output, measures the value of contributing content, and calculates proportional compensation based on a proprietary algorithm. It’s capable of handling text, images, music and video, and ProRata’s website says it’s available as a service for third-party generative AI systems. And this technology will sit at the heart of ProRata’s own gen AI search platform.
Other generative AI tools are compensating publishers too, mostly via wide-ranging licensing deals. But ProRata says its model, where compensation is based on each publisher’s contribution, offers a fairer deal. Unlike competitors, ProRata’s answer engine will only source content from licensed partners. While others allow publishers to opt-out of data harvesting, they’ll inevitably end up using a lot of third-party content for free from any publishers and creators who don’t have a licensing deal, but haven’t opted-out. (Some publishers also claim that their content is harvested even when they have opted out).
Publishers who have partnered with ProRata back its approach. “ProRata’s platform is a vital first step toward advancing accurate and fair attribution and promoting transparency,” said Rich Caccappolo, vice chairman of dmg media. “It could be the cornerstone of a sustainable economic model for news publishers, giving them the incentive to continue investing in high-quality, informative journalism.”
The test for ProRata and its compensation model will be whether it can generate similar enthusiasm from users once its answer engine launches.
