# Creator Discovery Moved Into AI Search. The Payment Didn't.

*Kosterina starts creator casting in Claude. Agencies are writing AI visibility into briefs. Creators are now cited inventory in two markets at once, and priced in neither.*

Published: 2026-08-18 | Read time: 5 min read | Author: Smalk AI Research | Source: https://www.smalk.ai/blog/creator-discovery-ai-search-payment-rail

---

Kosterina, an olive oil brand, now starts its hunt for creator partners inside Claude. That is a casting decision routed through an AI engine, which means creator discovery in AI search is not a content problem for creators. It is a market problem. Creators are now cited inventory in two markets at once, the consumer answer and the casting answer, and they are priced in neither.

- Brands and agencies have started shortlisting creators through AI chatbots, so a creator's citation footprint now shapes income, not just reach.
- Jellyfish's June 2026 analysis of 27 million AI response data points found YouTube creator content in more than one in four answers, and close to one in two in high-intent categories.
- The creators building answer-optimized personal websites are optimizing the smallest part of the surface, because citation is decided on third-party pages they do not own.
- Creator influence inside AI answers is rising while the price of that influence is still set by follower counts inherited from the social era.

## What creator discovery in AI search is

Creator discovery in AI search is the practice of brands and agencies using AI engines, rather than social platforms or influencer databases, to decide which creators to partner with. The prompt is the casting brief. The engine's cited sources are the shortlist. The creator never sees the selection happen and has no way to bid into it.

## The creator marketplace and the citation layer, side by side

In the platform economy, discovery ran through marketplaces both sides could see: follower counts, engagement rates, agency databases, rate cards. Value was measured, then priced. In the agentic economy, discovery runs through a synthesized answer that publishes no rate card and no inventory list. Influence is now measured, by a growing stack of visibility tools, but it is still not priced.

The mechanic changed. The underlying need did not. Brands still need to find credible voices, and the voices still need to be paid for being the source the engine leans on.

## What Digiday reported, and what the numbers add

Kimeko McCoy reported for Digiday on August 17, 2026 that creators are building their own visibility frameworks. Courtney Johnson has run a dedicated answer-engine-optimized site since October, with structured data and a plain-text index for machines; verified crawlers from OpenAI, Anthropic, Amazon, Apple, Huawei and DuckDuckGo read it, and it has produced 10,766 Google search impressions and 435 clicks since launch. Colin Rocker audited his own citations in Claude this month and found he surfaces mainly through his LinkedIn profile and podcast appearances.

One number needs correcting. Digiday cites a SponsorCX survey for the claim that nearly one in five consumers have replaced traditional research with AI. SponsorCX's own study page, based on 1,000 U.S. adults, reports 17% doing so regularly and another 34% occasionally, so 51% at least once. The fuller figure is the one that matters for planning.

## Why an answer-optimized creator site does not decide the citation

### The strongest version of the creator-leverage case

Creators genuinely are winning inside AI answers. Jellyfish, part of The Brandtech Group, analyzed more than 27 million AI response data points across seven LLMs, ten industries and fifteen-plus markets, and published on June 25, 2026 that YouTube creator content appears in more than one in four AI-generated answers, rising to almost one in two in high-intent sectors such as consumer electronics and financial services. In FMCG, nine out of ten YouTube videos surfacing in AI results came from independent creators rather than brands. Money is starting to move on that basis: Zoom is sponsoring a podcast series explicitly as part of an AI visibility strategy.

> It's not a direct correlation at this point,
>
> — Kimberly Storin, CMO, Zoom

### Why the most-cited voice still does not set the price

Zoom's own CMO concedes the link is unproven, and the deal was priced on pre-AI logic: audience, credibility, a speaking slot, with no citation measurement on either side of the table. That is a bet, not a market. Meanwhile the evidence says the deciding page is rarely your own property. Muck Rack's May 2026 study of 25 million citations put earned media at 84% of what ChatGPT, Claude and Gemini cite, with paid and advertorial content at 0.3%; Meltwater's April 2026 analysis of 5.35 million citations put earned and news media at 39.5%. The definitions differ, the direction does not.

Rocker's audit says the same thing at a sample size of one: LinkedIn and other people's podcasts, not his own site. And 435 clicks is roughly what self-optimization returns when there is no rail to sell the influence it creates.

## What this means for brands and for creators

### For CMOs, media buyers and agencies: brief for retrieval, not for reach

If your team is building shortlists in a chatbot, you are already buying citation influence, untracked and unpriced. Split the creator line in two: content bought for social reach, and content bought because engines retrieve it. Ask partners for a cost per mention alongside CPM, and assume competitors are querying the same engines with the same prompts on the same afternoon.

### For publishers and creators: sell the citation, not the follower count

Creators are the newest publisher class in the citation economy, and they inherit the publisher problem in compressed time: influence rising, compensation flat, terms set elsewhere. The answer-optimized site is table stakes, not the asset. The asset is documented presence on the pages engines actually consult, and the negotiating position is a citation footprint you can evidence, not an audience number that predicts nothing about retrieval.

## Three signals to watch before 2027

First, the measurement layer is forming fast, with Jellyfish adding Creator Intelligence to its Share of Model platform in June 2026 to identify which creators shape model recommendations. Second, agency briefs are starting to carry an AI visibility line, per Digiday's August 2026 reporting. Third, the format fight is live: Perplexity blocked Time's agent-directed ads in August 2026, calling them deceptive. Whoever standardizes payment to the cited source, publisher or creator, sets the price for everyone else.

## Conclusion

Hold on to this: creator discovery in AI search has quietly merged the audience market and the talent market into a single unpriced surface, and creators have joined publishers as suppliers of influence they cannot invoice. The structural fix is the same on both sides, an ad layer built for AI answers, which is the category Smalk AI is building: generative engine advertising that places native ads for AI agents on the sources engines rely on, and routes revenue to the media owners and creators feeding those answers. What to watch next: the first creator deal priced on a cost per mention instead of a CPM, because that is the moment citation stops being a bet and starts being inventory.

## FAQ

### What is creator discovery in AI search?

Creator discovery in AI search is when brands and agencies use AI engines such as ChatGPT or Claude, instead of social platforms or influencer databases, to identify creators for partnerships. The engine's cited sources become the shortlist, so a creator's visibility inside AI answers now influences commercial opportunity as much as follower count once did.

### Why do AI engines cite independent creators more than brands?

AI engines favor content that behaves like evidence: specific, comparative and explanatory. Jellyfish's June 2026 research found that in FMCG, nine out of ten YouTube videos appearing in AI results came from independent creators rather than brands, and that niche experts outperform both celebrity influencers and brand-owned content.

### Should creators build a dedicated AEO website?

Yes, but treat it as hygiene rather than strategy. Citation studies consistently show most AI citations land on third-party pages, so a personal site makes a creator machine-readable without deciding whether they get cited. The bigger lever is presence on podcasts, listicles, editorial coverage and video that engines already consult.

### How is paying a creator for AI visibility different from an influencer campaign?

An influencer campaign is priced on human attention: reach, engagement, CPM. Paying for AI visibility is priced on machine retrieval, meaning whether the content is cited when an engine answers a category question. Today most deals are still written on the first model while being justified by the second, which is why cost per mention matters.

### Do creators get paid when AI engines cite their content?

No. There is currently no standard mechanism that compensates a creator or publisher when an AI engine uses their content to build an answer. Compensation still arrives through brand deals negotiated on social metrics, which is the same gap publishers face between rising citation influence and falling monetization.