Quick Answer: Ads in Google AI Overviews appear as Search and Shopping ads inside or beneath the overview, under a "Sponsored" label. They serve from existing Search, Shopping and Performance Max campaigns. Google launched them for US mobile on 3 October 2024, added desktop and AI Mode on 21 May 2025, and Gemini-built AI Mode formats in May 2026.
Google has run ads inside AI Overviews for almost two years, and the format has changed more in 2026 than in the eighteen months before it. This page covers where the units sit, how they are labelled, what makes a campaign eligible, the new AI Mode formats, and the part most people get wrong: how the organic links in the same overview are chosen.
Everything below comes from Google's own announcements and its Search Central documentation. Where Google has not published a number, this page says so rather than guessing.
Where ads appear in AI Overviews and AI Mode
Ads appear in three places: above the AI Overview, below it, and, since October 2024, within the overview itself in a section labelled "Sponsored". Google says in-overview ads show when they are relevant to both the query and the response the overview gives, per its 3 October 2024 announcement.
In AI Mode, ads may appear below and integrated into the response. Google's example is a user asking how to build a website for a small business with limited resources, receiving a step-by-step answer, and then seeing a website builder's ad as a possible next step.
The rollout at a glance
| Date | What Google announced | Surface |
|---|---|---|
| 3 October 2024 | Ads in AI Overviews live for mobile users in the US after months of testing | AI Overviews, mobile |
| 21 May 2025 | Search and Shopping ads in AI Overviews expand to desktop in the US; English expansion to select countries later that year | AI Overviews, desktop |
| 21 May 2025 | Testing begins for ads in AI Mode, below and integrated into responses | AI Mode |
| January 2026 | Direct Offers pilot launches with brands including Chewy, Gap and L'Oreal | AI Mode |
| 20 May 2026 | Two Gemini-built formats announced for AI Mode: Conversational Discovery ads and Highlighted Answers | AI Mode |
How the units are labelled
Every unit carries Google's "Sponsored" label, in the overview, in AI Mode and in the 2026 formats. Google restated that commitment in its May 2026 post, saying the new formats "will also continue to be clearly labeled as Sponsored."
The label is the only visual separation a user gets. In an AI Overview the ad sits inside a block that reads like part of the answer, so the label is doing the work a separate ad column used to do on a classic results page.
Which campaigns are eligible to show in AI Overviews
Existing Search, Shopping and Performance Max campaigns are eligible, and Google says there is no additional action required. In its May 2025 update Google specified Performance Max, Shopping, and Search campaigns using broad match, including AI Max for Search, per its Google Marketing Live post.
What that means in practice
There is no AI Overviews campaign type to create and no placement to opt into. If a Shopping campaign is live and the query triggers an overview where the product is relevant to the response, the ad is eligible. The lever you control is campaign type and match type, not a checkbox.
What you cannot control
Google's announcements describe no AI Overviews placement report, no separate bid, and no way to exclude the overview while staying in classic results. The AI features are treated as part of Search, so eligibility travels with the campaign rather than being bought separately.
How the 2026 AI Mode ad formats work
The two formats Google announced on 20 May 2026 are built with Gemini and each carries an independent AI explainer alongside the advertiser's creative. Google says its model "evaluates and synthesizes information about a product or service, and displays that context alongside the advertiser's creative," per the announcement.
Conversational Discovery ads
A Conversational Discovery ad answers the person's specific question. Google's example is a query about low-maintenance ways to make a home smell like a spa, where Gemini builds ad creative tailored to that search and highlights the relevant features of the product.
Highlighted Answers
A Highlighted Answer is an ad placed on a list of recommendations that AI Mode produces. Google's example is a list of the best language apps for a trip, where a highly relevant, high-quality ad becomes eligible to appear on that list. This is the format closest to an organic recommendation, which is why the Sponsored label and the independent explainer matter.
Direct Offers and the formats coming to Search
The Direct Offers pilot, running since January 2026, lets brands surface deals as shoppers explore their options. Google is adding promotion bundling, native checkout for Universal Commerce Protocol merchants, and travel partners including Booking and Expedia. Two further formats, AI-powered Shopping ads and Business Agent for Leads, are announced for classic Search in the coming months.
How organic AI Overview links are chosen
The organic links in an AI Overview come from Google Search's own systems, with no additional requirements beyond fundamental SEO. Google states that directly, in its AI features documentation, and the ads system decides the sponsored unit separately.
Why the two systems are separate
Google describes an in-overview ad as shown when it is relevant to both the query and the response. That wording implies the response is generated first and the ad is matched to it afterwards. The links inside that response are identified by what Google calls query fan-out: multiple related searches across subtopics, with supporting pages selected as the response is generated.
That order matters for a brand. A page can be one of the overview's cited links without any ad spend, and an ad can sit beneath an overview whose links go to a competitor. Ranking in Google AI Overviews is a content and crawl question; the ad is a campaign question.
What Google says you do not need
Google's documentation says no new machine-readable files, AI text files or markup are required to appear in AI features, and no special structured data either. The same robots.txt directives for Googlebot govern crawling for Search and its AI features, so robots.txt rules for AI bots that block Googlebot remove you from both.
What moves the organic slot
The same fundamentals that rank a page: crawlable, indexable, answering the question directly, with a claim the model can attribute. Structuring content so an AI can extract it is the practical version of Google's advice, and it applies equally to AI Mode, which may use different models and show a different set of links.
How to measure ads and organic in the same overview
Measure the paid unit in Google Ads and the organic link in Search Console, because Google does not report the two together. Google's documentation says sites appearing in AI Overviews and AI Mode are included in overall search traffic in Search Console, in the Performance report under the Web search type, with no separate filter.
The paid side
Ads eligible for AI Overviews and AI Mode report inside their existing campaigns. Google has not announced an AI Overviews column, so the practical method is to compare campaign performance on queries that trigger overviews with queries that do not, using the query report and a manual check of which ones show an overview.
The organic side
Read the overview from a signed-out search in your target country and record three things per query. Note whether an overview appears, whether your page is one of its links, and whether your brand is named in the answer text. Measuring AI visibility on a fixed schedule turns that into a trend, and tracking AI traffic in Search Console and GA4 attaches the click.
The one comparison worth making
Put the paid cost per click for an overview-triggering query beside the organic clicks the same query sends from the overview link. Where the organic link is already yours, the ad is defending a slot you hold; where it is not, the ad is renting attention on an answer that recommends someone else, and answer engine optimization is the cheaper fix.