How Will Yahoo Scout Pay Publishers Through the Internet Licensing Collective?

What Exactly Launched on November 12?

A licensing body, a first customer and a price estimate, all on the same day. The ILC describes itself as “a nonprofit, publisher-led collective rights organization making AI content licensing practical at internet scale.” It was previously called the RSL Collective, after the Really Simple Licensing (RSL) standard it co-wrote. RSL is a machine-readable way for a website to state the terms on which AI companies may use its content (ILC release).

The first customer is Yahoo Scout. Eric Feng, senior vice president and general manager of Yahoo Research Group, said in the release: “AI answers are only as good as the sources behind them.” When we requested scout.com on November 13, it redirected to Yahoo Search in AI mode (search.yahoo.com/?mode=ai). Search Engine Roundtable reported the deal on November 13 and noted that no payment figures were given (Search Engine Roundtable).

The third piece is a report. Mather Economics, a publishing analytics firm, released The Price of Intelligence the same day. The ILC says Mather started the study on its own and that it “was not commissioned or funded by ILC” (ILC release).

How Does the Money Reach a Publisher?

Through one pool, split by use. An AI company takes one blanket license and picks the uses it needs: training, retrieval-augmented generation, grounding, agentic applications or AI-generated overviews. The ILC handles the repertoire, verification, reporting and payouts “at cost.” After costs, which it expects to be 10% or less, the rest goes to rights holders (ILC release).

The split follows three factors listed on the ILC’s homepage (ILC):

• Contribution. Work that shapes an answer, grounds it or is cited as a source counts for more.
• Frequency. Content used across many outputs is worth more than a one-off.
• Value of the AI use. Higher-value AI products feed larger revenue pools.

The FAQ says publishers are paid “on a regular schedule” and can see licensee-reported usage on a dashboard, by website. It gives no payment frequency, no minimum payout and no description of how usage reports are audited (ILC FAQ).

How Much Could It Pay?

Nobody has said what Yahoo pays. The only figures are Mather’s estimates for the whole market. Mather says AI companies have “an economic willingness to pay between 15% and 35% of revenue” for licensed content. It puts collective licensing “in the tens of billions,” based only on the run-rate revenue of the two biggest frontier models (ILC release; Mather Economics).

We checked the ILC homepage figures against that range. The site gives 2030 estimates of “$500B+” in AI revenue and “$75–170B” in licensing (ILC). By our arithmetic, $75 billion is 15% of $500 billion and $170 billion is 34%. So the homepage range is Mather’s percentage range applied to one revenue forecast. It is not a separate estimate.

Our arithmetic on the cost cap: if operating costs come in at the projected 10%, every $100 of license fees leaves at least $90 for rights holders. If an AI company paid at the bottom of Mather’s range, 15% of revenue, $1 million of product revenue would put $135,000 in the publisher pool. At the top, 35%, it would be $315,000. These are illustrations of the stated percentages, not terms of the Yahoo deal.

The chart shows how far the estimate sits below the music and video benchmarks the ILC cites: even the top of Mather’s range is half of what the release says Spotify pays.

Are the Featured Publishers Set Up to Be Paid?

On their main sites, mostly not yet, by our check. The ILC’s publisher guide sets out the technical steps. A publisher posts an RSL file whose payment terms point to https://internetcollective.org/license, then adds a line to robots.txt. Each licensed host “must make its RSL files discoverable through robots.txt by adding one or more License: lines.” After that, the publisher enrolls each website in the ILC portal (ILC Publisher Guide; RSL 1.0 specification).

On November 13, 2026, between 8:30 and 9:00 a.m. Eastern, we fetched the robots.txt file of the main site of each of the 13 brands named on the ILC homepage. We looked for License: lines, Content Signals lines (a newer robots.txt line for stating AI preferences) and full blocks on three AI crawlers.

Three files carry a License: line. We could read two of the files those lines point to, and neither references the ILC license. The Guardian’s file permits ai-train and ai-input under a subscription term that links to its own licensing site (Guardian license.xml). Stack Overflow’s points to the Creative Commons BY-SA 4.0 license (Stack Overflow license.xml). Medium’s file sat behind a Cloudflare challenge we did not get past. Stack Overflow answered our robots.txt request with HTTP 418 and a file that disallows every crawler, which may be a response to automated visitors, so we do not report its crawler rules. The Atlantic is the only featured brand with a Content Signals line: search=yes, ai-input=no, ai-train=no (The Atlantic robots.txt).

The wider sample tells the same story. We ran the same check on 20 other large news and tech sites: The Washington Post, CNN, BBC, Bloomberg, Reuters, NBC News, Fox News, NPR, the Los Angeles Times, BuzzFeed, Vox, The Verge, Wired, Forbes, Business Insider, TechCrunch, CNET, Mashable, Quora and Yahoo. None carries a License: line. Among them, 17 block ClaudeBot, 12 block GPTBot and 9 block Google-Extended. Yahoo’s own robots.txt blocks GPTBot, ClaudeBot, Google-Extended, PerplexityBot and CCBot (Yahoo robots.txt). In September our check of 15 news sites found a similar pattern of crawler blocks, covered in Do News Sites Block AI Crawlers?

Why Would a Member Not Have the Line?

The guide allows other routes, and the launch is one day old. The publisher guide also lists on-page declarations, HTTP headers, RSS feeds and private access URLs. A publisher can license one section or subdomain, not the whole site. A publisher can also back the RSL standard without enrolling in the collective’s license. The release says RSL launched in 2025 “with the support of more than 1500 internet publishers and technology companies, including Yahoo” (ILC release). Supporting a standard and licensing through the collective are different steps.

That matters because a robots.txt line is public and checkable, and a portal enrollment is not. Licensing deals in this market have mostly kept their terms private, as our report on which publishers have AI licensing deals found. Under the ILC, anyone reading a robots.txt file could in principle confirm a publisher’s participation.

What This Means for Anyone Buying a Placement

A placement on an ILC member site does not buy an AI citation. The allocation rules reward publishers whose work AI products use. They give advertisers, contributors and the companies being covered nothing. Licensing also does not mean a site lets AI crawlers in: 10 of the 12 featured sites whose crawler rules we report block at least one of the three crawlers we checked. If you want to know whether a crawler can read a page, check the robots.txt rules yourself; our guide to blocking or allowing AI crawlers in robots.txt explains how to read them.

Our recommendation, not a rule: when a publisher says its coverage reaches AI answers through licensing, ask which licensees it covers and whether its robots.txt carries a License: line. Our earlier explainer, Can You Charge an AI Crawler?, covers the pay-per-crawl alternative.

Methodology and Limitations

Four limits apply. First, we checked only the main host of each brand, following redirects. A brand may have licensed a subdomain, a section or a sister site we did not check, and may use the RSS, header or on-page methods the ILC allows. Second, robots.txt shows what a site says at one moment, from one vantage point. Third, “blocked” means a Disallow: / rule for that crawler (for Reddit, for all crawlers); path exceptions are not shown. Fourth, Mather’s figures come from the ILC release and Mather’s summary page; we did not review the full PDF. The next thing to watch is whether the ILC publishes its Allocation Rules in full and names a second licensee.

Sources and Methodology

This explainer is built from the ILC’s November 12, 2026 launch release, homepage, FAQ and publisher guide, the RSL 1.0 specification, Mather Economics’ summary page for The Price of Intelligence, Search Engine Roundtable’s November 13 report, and the live robots.txt and RSL files of 33 websites, all read November 13, 2026. The robots.txt table, the 20-site comparison counts, the check of the ILC homepage range and the payout illustrations are our own work. Advice to buyers is our recommendation, not a rule. We did not contact the ILC or Yahoo.

1. official Internet Licensing Collective, The Internet Licensing Collective (ILC) Launches with Yahoo Scout as First Commercial Partner (November 12, 2026). Launch, membership, license uses, allocation, cost cap, quotes, Mather findings. Read November 13, 2026.
2. official Internet Licensing Collective, Homepage. 3,100+ members, 3.1B+ works, allocation factors, 2030 estimates, featured brands. Read November 13, 2026.
3. official Internet Licensing Collective, FAQ. Fees, exclusivity, payment schedule, dashboard, pricing for AI companies. Read November 13, 2026.
4. official Internet Licensing Collective, Publisher Guide: onboarding. RSL file syntax, robots.txt License line requirement, enrollment. Read November 13, 2026.
5. standards RSL Standard, RSL 1.0 specification. License, usage and payment elements. Read November 13, 2026.
6. vendor research Mather Economics, The Price of Intelligence (summary page). 15% to 35% willingness to pay, method. Read November 13, 2026.
7. trade press Search Engine Roundtable, Yahoo Scout To Pay Publishers Through The ILC For Using Content In AI Answers (November 13, 2026). Read November 13, 2026.
8. live read robots.txt files of 33 sites, including The Guardian, The Atlantic and Yahoo, plus the license files of The Guardian and Stack Overflow. Read November 13, 2026.