This conversation is for our exclusive publishers and partners only community. Please ask in the community for the password to watch this video.

The shift is already here. When AI summaries and answer engines sit between users and source sites, clicks fall, sessions shrink, and the value exchange around publisher content erodes. In October’s Camp.Fire session, rev ops leader Scott Messer framed the moment clearly:

“This is the crawling to collaboration phase.”

This is a transition most publishers still have not operationalized. So where do publishers start?

1. Build leverage together

Scott presented data points that match what many publishers are seeing internally: AI overviews cut into engagement, and chat results send far fewer clicks than the old-school search page ever did. The exact percentages may vary, but the trend is clear, and it’s not upward.

If your content is fueling third-party models and experiences, i.e. someone else’s AI, you deserve a seat at the table. Right now, one-off licensing deals are like a chain-link fence trying to stop a flood of crawlers and scrapers. Too many deals, all on different terms, just don’t scale.

As Messer put it:

“Stop waiting for big tech to offer you a fair deal, because guess what? Big tech doesn’t do fair deals.”

Several routes were discussed. A legal path could eventually compel licensing, but that’ll take years. Meanwhile, a unilateral fix from a single platform is possible, yet unlikely to arrive on publisher-friendly terms. Which leaves the publisher-led path: align on practical frameworks, create real friction for non-compliant access, and provide clear avenues to transact.

No single publisher can create that alone, so collaboration and participation are crucial.

2. Escape from the “panic committee”

Scott described the four stages, as he sees it, of where publishers are in their AI licensing process right now:

1. Head in the sand: The company is hoping it blows over, or decision-makers aren’t paying attention.
2. Panic committee: There’s technically a group, but no-one owns it, meetings get canceled, and everyone’s just sharing links instead of taking action. And the CEO may not be aware of it either.
3. Group soup: Some momentum exists, opinions are forming, and learning is happening, yet the team still cannot take decisive action.
4. Action with ownership: There’s a clear owner and a mandate to act. The company has a seat at the table for licensing and negotiation, has reached out to vendors, is engaging with frameworks, and is participating in working groups.

A lot of teams are still parked in stage one or two. If that sounds familiar, it’s time to pick an owner, give them real authority, and actually move on your AI plan.

3. Get specific about standards and enforcement

A policy without the plumbing behind it will not move the needle. Start by tightening access. Audit your sites’ robots.txt (instructions to web crawlers) and permissions, then pair policy with instrumentation so you can see who is visiting, what they take, and how deep they go. Scott described using bot analytics to track what model builders are doing, turning a worry into a weekly metric you can actually act on.

From there, you can decide where to block, throttle, or route to paid paths. However, don’t assume that a cosmetic fix solves a platform risk. As Scott put it:

“Don't think that because you're blocking Google Extended, the Gemini crawler, that you’ve solved your Google problem.”

Engage in standards work that is actively building the pipes for consent and compensation. For example, the IAB Tech Lab’s Content Monetization Policy (CoMP) working group is meeting regularly and exploring schemas that look familiar to programmatic practitioners, so get involved. Send more than one person, including an engineer and a product manager, so you can influence the spec and prepare your stack to implement it.

4. Treat it like a product – narrow the options and build a market

Markets only take off when buyers can understand, compare, and trust a few clear options. If publishers all show up with different terms, AI companies will just write their own playbook. Your job is to help collapse the menu into a few interoperable methods, each supported by multiple vendors. Make sure they have clear discovery, pricing, and measurement. That’s how programmatic matured, and how licensing for AI use could mature in the not-so-distant future.

Inside your company, run like a product team. Hold “lunch and learn” sessions across departments to get everyone - from editorial to ad ops - on the same page. Shortlist a few licensing frameworks and vendors, take demos, and give them product feedback. Above all, bring an opinion to the table, then evolve it. Sit back, and you’ll live with someone else’s standard. Get involved, and you’ll help shape it.

5. Now, name your AI expert and build a coalition

The work starts with a simple question: who at your company owns AI licensing and clickless distribution strategy today? If the answer is no-one, then volunteer to draft the first plan, and recruit the right person to drive it.

Rob wrapped it up with a useful reminder:

“Don’t think just putting AI on something solves it. We have a lot of work to do at very basic, fundamental levels.”

If you align internally, show up externally, and help reduce chaos into a few workable paths, you’ll be ready before the clickless curve gets any steeper.