The meter goes live
SPUR, the Standards for Publisher Usage Rights initiative, published version one of its content telemetry standard today. Formed in March by a coalition that includes the Guardian, the Financial Times, the BBC, Sky, the Times of London, MediaHaus and the AP, the group put out a draft in June, took public comment through July 24, and has now shipped the first version.
The standard creates a process for tracking what happens when publisher content meets an AI tool: when it is retrieved, when it is grounded (used as the underlying evidence for an AI answer), when it is cited, when it is presented, and when it is engaged with. Those signals get reported back to the publishers. This version also handles multimodal content, not just text, and is designed to plug into existing provenance work like C2PA.
Data before deals
The logic is worth spelling out. The training-data lawsuits never produced a price list. You cannot license what you cannot measure, and right now neither side has a shared record of how much AI answers lean on publisher reporting.
Alex Springer, SPUR's technical lead, described the goal as assembling a stack that is easy for an agent to use: licensing payments where they apply, free content where it fits, reporting, provenance. The whole chain. Springer also made the honest case that licensing is the slow path: it is trivial for a company to pull free API credits from a grounding service and start using content today, while doing it properly takes time and effort. A standard that makes the proper path easy is, at least, removing the excuse.
The catch: the labs are building their own meters
A standard only matters if the other side adopts it. SPUR has invited OpenAI, Anthropic, Google, Meta and Microsoft to shape implementation through a new invitation-only AI Licensing Advisory Board, expected to hold its first meeting this month. Responses so far tell you everything about where this stands: Google offered a non-answer about frequently engaging with SPUR and other associations, and the other companies did not respond before publication.
Meanwhile Microsoft, Google and OpenAI have started releasing their own implementations of content usage reporting. Springer pointed to retrieval and grounding events appearing in Google and Microsoft systems, and put it plainly: that action comes from somewhere. SPUR is also building agent tooling to make adoption easier, and says pilot programs with tech and AI companies are next, though it declined to name who takes part.
What is next
On October 15, SPUR's first technical committee meets to work out the framework's "auditing and evidencing": how to validate that the signals AI tools send back to publishers are real. That is the right thing to nail down first, because a meter nobody trusts is just decoration.
The deeper shift is one of posture. Publishers are drifting toward block-by-default against AI crawlers. Springer argued that data exchange is the thing that keeps them open: a reason not to block, in exchange for visibility into how agents use their work. If metering works, the AI answer economy gets a functioning licensing layer. If it does not, the lawsuits keep going and the blocklists keep growing. Michael Rubenstein of Firsthand, an AI brand agent platform, framed the stakes as journalism's survival: publishers need a sustainable revenue stream to keep investing in reporting, because an internet of degraded content produces degraded answers.
Sources
- [1] Digiday, October 2, 2026, “SPUR publishes AI tracking standard, invites OpenAI and Google” by Sara GuaglioneRead source