A consortium of media organizations, which includes prominent names like the Guardian and the Financial Times, has introduced a new standard aimed at monitoring how artificial intelligence (AI) tools utilize publishers’ content. The initiative, named Standards for Publisher Usage Rights (SPUR), unveiled its content telemetry standard today and has reached out to major AI companies such as OpenAI, Anthropic, Google, Meta, and Microsoft to collaborate on its implementation, according to Alex Springer, SPUR's technical lead.
If these companies choose to accept the invitation, they would be part of SPUR’s exclusive AI Licensing Advisory Board, which is designed to ensure the standard encompasses a comprehensive view of the entire AI landscape. When asked about Google's participation, a spokesperson stated, “We frequently engage with SPUR and other associations on a variety of topics.” The remaining companies have not provided any comments prior to the publication.
Established in March, SPUR also counts the BBC, Sky, The Times of London, European media group MediaHaus, and the Associated Press among its members. The newly released standard establishes a framework for tracking and reporting the instances when AI tools retrieve, ground (that is, use as reference material), cite, present, and engage with content, subsequently relaying that information to the publishers.
The board aims to create tracking protocols that benefit both publishers and those developing AI technologies. If widely adopted, this standard could provide publishers with reliable data regarding when their content is accessed, cited, or featured in AI-generated responses, valuable insights that could inform licensing negotiations, as noted by Springer.
The advisory board is expected to comprise approximately 20 members, with its first meeting scheduled for October.
David Buttle, co-founder of SPUR, emphasized through an email to Digiday that the initiative ensures all parties have a say, encouraging cooperation toward more equitable and sustainable practices.
Springer pointed out that it's relatively easy for companies to obtain free API credits from services that offer grounding or scraping tools like Exa and Parallel, allowing them to access content without proper licensing, which typically requires more time and effort.
“The aim is to integrate this into a coherent suite of solutions that agents can easily utilize,” Springer explained. “This involves licensing payments — and while free content definitely has a role, there is also the need for accurate reporting of its origins.”
After releasing a draft of the content telemetry standard in June, which was open for public feedback until July 24, the coalition has since developed its initial version. Notably, the first version now accommodates multimodal data rather than being restricted to text-based content, and it has been refined to better align with other technical standards, such as C2PA and provenance protocols.
The coalition is actively collaborating with tech and AI firms to execute the standard and is preparing to initiate pilot programs with them. While Springer refrained from naming participating companies, he acknowledged that firms like Microsoft, Google, and OpenAI have already begun to implement their own content usage reporting mechanisms.
“It’s not the case that OpenAI has commented specifically on our standard. They are aware of it... However, when you start seeing retrieval and grounding events in the systems of Google and Microsoft, it indicates there’s a source for that action,” Springer mentioned.
In addition to this, SPUR is developing tools that will assist AI companies in adopting the standard, thereby promoting more transparent reporting and licensing practices.
Michael Rubenstein, co-founder and co-CEO of Firsthand, an AI brand agent platform that crafts adaptive advertising experiences, remarked, “A better internet hinges on publishers being able to generate sustainable revenue, allowing them to continue supporting high-quality content and journalism. A decline in content quality would raise concerns, especially if answer engines increasingly rely on such degraded information.”
On October 15, SPUR’s inaugural technical committee will convene to establish the framework for “auditing and evidencing,” focusing on methods to validate signals sent from AI tools to publishers.
“It’s about more than just publishers wanting to keep count for billing purposes,” Springer clarified. “It’s about understanding what is valuable, how to amplify that, and how to pinpoint necessary inputs for engagement.”
He argued that such an approach could incentivize AI companies to engage more directly with publishers, offering them better insights into how their content is being accessed and utilized, which might discourage publishers from outright blocking access.
“Currently, the tendency among publishers is to block. We are moving towards a default ‘just-block’ approach, but the potential for data sharing and engagement can help keep that channel open,” Springer concluded.
Note: An earlier draft of this article inaccurately suggested that Google and OpenAI had been approached regarding the standard; it has been corrected to clarify their invitation to join the Advisory Board.
