Skip to main content
SEO InsightsAI Search

Ads Are Moving Inside the AI Answer. Here Is What That Means for Your Visibility

Summarize with ChatGPT
JK
John Kyprianou
June 24, 2026
7 min read
An AI search answer with a sponsored result woven into the response text instead of sitting above it

The line between an ad and an answer just got blurry on purpose.

For 25 years the deal was simple. Paid results sat in a labelled block at the top, organic links sat below, and a user could tell the difference at a glance. In June 2026 both Google and OpenAI started dismantling that arrangement. Ads are no longer stacked above the answer. They are being written into it.

An AI search answer with a sponsored result woven into the response text instead of sitting above it

At Google Marketing Live, Google introduced a set of ad formats built with Gemini that appear directly inside AI Mode. ChatGPT, in the same month, opened conversion-optimized campaigns to advertisers. Two of the biggest answer engines on the planet decided, within weeks of each other, that the answer itself is now ad inventory.

Here is our read on what changed, what it means for the unpaid side of search, and what we would actually do if it were our budget.

What actually changed in June 2026

Google announced four Gemini-built ad formats, and two of them live inside the AI answer rather than above it.

Conversational Discovery ads read the specific question a person asks and have Gemini assemble creative that highlights the product features relevant to that exact query. Highlighted Answers let a "highly relevant, high-quality ad" slot into the recommendation list AI Mode generates, sitting alongside the organic suggestions. Google also showed AI-powered Shopping ads that write a custom explainer for high-consideration buys, and a Business Agent for Leads that drops a chat agent trained on your site straight into the ad (Google Ads blog).

OpenAI moved in parallel. Conversion-optimized ChatGPT campaigns began rolling out in early June, adding a Conversions objective alongside Reach and Clicks, with an OpenAI Pixel to track post-click activity and a Conversions API to feed first-party data back in. These run through the ChatGPT Ads Manager, OpenAI's self-serve platform (Search Engine Roundtable).

Strip away the product names and the shift is the same on both platforms. The answer is no longer a neutral zone. It is becoming the most valuable square of real estate in search, and it is now partly for sale.

The mechanics changed too, and that part matters more

The bigger story is not where the ad sits. It is how the ad gets built.

Classic search ads were keyword-driven. You picked the words that triggered your ad and wrote the copy yourself. The new formats do not work like that. You hand the system inputs, your product feed, conversion data, audience signals, and your landing page content, and it assembles the ad to fit the question. You no longer choose the exact words. You choose the raw material.

That is the detail most people are skating over. The inputs that fuel these AI ads are the same inputs that decide whether you get cited organically. Clear product information, well-structured pages, real proof, clean feeds. The machine reads your site the same way whether it is deciding to quote you for free or build a paid unit around you.

So the convergence everyone keeps predicting between organic and paid is not coming. It quietly arrived in the plumbing.

Why this makes organic citations more valuable, not less

The obvious reaction is fear. If ads are inside the answer, organic visibility must be worth less. We think that is backwards.

When the answer becomes monetized, an unpaid citation inside it becomes the scarcer and more trusted thing. Users are already learning to spot the "Sponsored" label, and Google has been careful to keep these formats clearly marked. The recommendation that the model reaches for on its own, with no ad spend behind it, carries a credibility that a labelled ad cannot buy.

There is a second reason. Ad inventory inside an answer is finite. Only so many sponsored slots can appear before the answer stops being an answer. Organic citation has no such cap. If the model trusts your content, it can pull from you across thousands of long-tail questions that will never be worth anyone's ad budget.

We see this constantly with client work. The pages that earn citations are not the ones chasing a single high-value keyword. They are the ones that answer a whole cluster of real questions cleanly enough that a model can lift a sentence and trust it. That is the asset ads cannot replace.

The risk nobody is pricing in

Here is the part that should make you cautious before you move budget.

Buying your way into the answer with ads is now easy. Buying your way into the organic citation is still against the rules, and Google has said so plainly. In June it confirmed that its search spam policies apply to AI search features, and it specifically warned about manipulating or buying citations. We wrote about that distinction in can you buy AI citations, and the line has only got sharper since.

So the two paths into the answer are not symmetrical. One is a transparent auction. The other is earned, and the moment you try to shortcut it you are in spam territory. Treat them as different disciplines with different rules, because Google does.

What we would do right now

Do not panic-shift your whole budget into AI ads, and do not ignore them either. Both reactions are lazy. Here is the order we would work in.

Get citable first. The content that earns organic mentions is the same content that makes your paid units relevant. Tight topic clusters, direct answers near the top of the page, named authors, real proof. This is the work that compounds, and it feeds both sides of the answer. Our take on structuring content so LLMs cite you still holds.

Then test the new ad formats with proper tracking. If you run paid, the conversion objectives on ChatGPT and the AI Mode formats on Google are worth a controlled test. The key word is controlled. Set up the Pixel or Conversions API, measure actual conversions, and do not let "we are inside the AI answer" become a vanity metric. A sponsored slot you cannot attribute is just spend.

Keep your feeds and landing pages clean. Because the system builds the ad from your inputs, a messy product feed or a thin landing page now hurts you in paid as well as organic. The housekeeping that used to be a technical SEO chore is now also an ad-quality lever.

Watch your own answer share. Track how often you appear inside AI answers, paid and unpaid, for the questions that matter to your business. This is the new version of rank tracking, and it is where attention should go.

The bottom line

Ads moving inside the AI answer is not the end of organic search. It is a repricing of it.

The answer is now the prize, and there are two ways into it. One you can buy, transparently and at a cost that will rise. One you earn, slowly, and that no competitor can outbid you for. Smart businesses will run both, but they will not confuse them, and they will keep investing in the earned side precisely because the paid side is getting crowded.

If you are weighing how to split effort between getting cited and paying to appear, that is exactly the kind of call we help with. Our AI search optimization and PPC work sit on the same strategy table for this reason, and you can start with a free SEO review to see where you stand inside the answer today.

John Kyprianou

John Kyprianou

Founder & SEO Strategist

John brings over a decade of experience in SEO and digital marketing. With expertise in technical SEO, content strategy, and data analytics, he helps businesses achieve sustainable growth through search.

Related Articles

Split diagram showing ChatGPT pulling English pages while Google AI Overviews pulls local language pages, by SEO Turtle
AI Search

ChatGPT Prefers English. Google's AI Prefers Your Customer's Language.

Two studies published this summer found the same split from opposite directions: ChatGPT over-retrieves English pages, and Google AI Overviews almost never leaves the language of the question. If you run a Greek and English site, or English and Spanish, you have two different blind spots and one bad reflex to avoid.

August 7, 2026
Diagram of a buyer leaving an AI answer to verify it on Reddit, reviews and search, by SEO Turtle
AI Search

AI Search Trust Is Falling While Use Keeps Rising. The Gap Is Where You Win

People are using AI search more and trusting it less. New research puts the drop at 28 points in a year, and half of US shoppers now fact-check AI product recommendations on Reddit before buying. The answer box is no longer the finish line. Here is what to put in the verification step.

August 5, 2026
Diagram showing an AI answer citing a source link without naming the brand behind it, by SEO Turtle
AI Search

Ghost Citations: When AI Cites Your Page But Never Names Your Brand

AI engines are linking to your content without ever saying who wrote it. New research puts the rate at around 40 percent of citations, and above half on Perplexity. That gap breaks the way most businesses measure AI visibility, and it changes what you should be writing.

July 31, 2026
A short three word keyword search above a much longer planning query full of context and constraints, illustrated by SEO Turtle
AI Search

Planning Queries Are the Fastest-Growing Search Behaviour, and Almost Nobody Has a Page for Them

Google published data on how people actually use AI Mode, and the fastest-growing query type is not what SEO teams have been building for. Planning and brainstorming searches are pulling ahead, and the average business site has nothing that answers them. Here is what those queries look like and what to build.

July 29, 2026
Google AI Overviews generating an image from a text prompt instead of showing a photo from a website, illustrated for Cyprus and US businesses by SEO Turtle
AI Search

Google Is Generating Images Inside AI Overviews. Now What?

On the 25th birthday of Google Images, Google added image generation to AI Overviews. Search can now draw the picture instead of showing yours, with no photographer, no attribution and no click. Here is what actually changes for image SEO, and the part of visual search a model still cannot fake.

July 27, 2026

Continue Your SEO Journey

Explore more expert insights and take action on your SEO strategy