Ads in AI Answers Tracker: Where Sponsored Content Appears in AI Search

Advertising has moved inside AI answers, and each provider places, labels and targets those ads under different rules. Ads in AI answers are paid placements shown within or beside a response written by an AI search or chat product, such as a sponsored product list in Google AI Mode or a labelled ad at the bottom of a ChatGPT reply. This tracker records, from each provider's own announcements and help pages, which products carry ads, where they appear, how they are labelled, which markets they reach, what advertisers control and what users can switch off. It is written for marketers and advertisers planning budgets, publishers watching how answers are monetised, privacy and security teams writing acceptable-use rules, and anyone who wants dated facts rather than commentary. Providers are listed alphabetically; order is never a ranking.

Last checked: 2026-11-15.

This is an independent reference page from NHI Mgmt Group based on public provider documentation; it is not sponsored or endorsed by any provider.

Key takeaways

• Three large providers have published ad programmes for AI answers. Google (AI Overviews and AI Mode), Microsoft (Copilot) and OpenAI (ChatGPT) have each announced ads in or beside AI responses, with dates and markets set out below.
• Every provider states that ads do not change the answer. Google, Microsoft, OpenAI and Perplexity all say in their announcements that sponsored content does not influence organic AI responses.
• Labelling is consistent in principle, different in wording. Google labels ads "Sponsored", Perplexity used "sponsored" for its 2024 test, OpenAI says ads are clearly labelled and separated, and Microsoft says Copilot ads are distinctly marked.
• Advertisers often cannot target AI surfaces on their own. Google's help centre says existing Search, Shopping and Performance Max campaigns are eligible automatically and that advertisers cannot opt out of AI placements separately from Search.
• Ad-free options exist and are stated plainly. OpenAI says Plus, Pro, Business and Enterprise plans do not include ads, and Kagi states it has no ads and is funded by its users.

Which AI answer products show ads?

The table gives one row per product with identical fields. Each status is taken from the provider's most recent official page found for this tracker, with that page's date. Where a provider's page does not state a field, the cell says so.

Table as of 15 November 2026.

Perplexity status note: the 12 November 2024 post is the latest Perplexity advertising announcement on its own blog found for this tracker. Search Engine Land reported in 2026, citing the Financial Times and unnamed Perplexity executives, that the company had ended the ad tests it began in 2024. Treat the Perplexity row as historic until Perplexity publishes its own update.

For how each engine finds and cites sources in the first place, see the AI Answer Engines Hub.

Where do ads appear and how are they labelled?

Placement and labelling are the two facts users and regulators look at first. The wording below follows each provider's own page; quotation marks mark the provider's words.

Two placement details matter for anyone auditing what users see. First, Google's help centre separates ads inside the AI response from ads above or below an AI Overview: the second type is a normal search ad on the same results page. Second, Google states that AI explainers are not part of ad headlines or descriptions, and that advertisers cannot write, edit, approve or opt out of them; advertisers can only keep their website content clear and current.

Which countries and languages are covered?

Market lists change faster than any other field on this page. The rows below quote the official pages as of their dates; re-check the source before planning spend.

Table as of 15 November 2026.

Google adds that its country list "may not be exhaustive" and is expanding over time. The October 2026 blog post names the ten markets added in that release (Australia, Canada, India, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, the Philippines and Singapore) and launches multiturn ads in the US so that ads can appear in follow-up queries in AI Mode.

What do advertisers control?

For buyers, the practical question is whether AI placements can be bought, excluded or measured separately. The answers differ by provider.

ChatGPT (OpenAI)

• Buying: the 16 January 2026 announcement describes the test from the user side and does not describe advertiser buying tools.
• Landing pages: OpenAI's crawler documentation lists OAI-AdsBot, which "only visits pages submitted as ads" to check landing-page safety; OpenAI states the data it collects is not used to train generative AI foundation models, and that landing-page content may be used to decide when an ad is most relevant.
• Sensitive topics: ads are not eligible near sensitive or regulated topics, named as health, mental health and politics.

Google AI Overviews and AI Mode

• Buying: existing Search, Shopping and Performance Max campaigns are eligible automatically, with no new campaign or budget required.
• Targeting and exclusion: advertisers cannot target AI Overviews or AI Mode exclusively, and cannot opt out of them separately from Search. Negative keywords, brand exclusions and guardrails such as asset pinning still apply.
• Matching: ads are matched directly to queries with clear commercial intent, or contextually to the query and the AI response for exploratory questions. Exact and phrase match keywords are not eligible for contextual matching on broader exploratory questions.
• Pricing: the same auction and cost-per-click or conversion-based bidding as standard Search ads, with no extra charge to participate.
• Reporting: metrics are included in aggregate in standard campaign reports, and triggering queries appear in aggregate in the search terms report.
• Policy: all Google Ads policies apply.

Microsoft Copilot

• Buying: the March 2025 post describes Showroom ads and dynamic filters as pilots with select clients, and a pilot that lets brands build AI agents on their own websites.
• Relevance: Microsoft says Copilot ads appear only when highly relevant and from trusted sources.
• Advertiser tools: Microsoft's April 2026 post covers Copilot inside the Microsoft Advertising Platform (performance shift root cause analysis and conversion tracking diagnostics) and AI Max for Search, in pilot globally excluding China. These are advertiser tools, not placements inside Copilot answers.

Perplexity

• Answers to sponsored questions: Perplexity's 2024 post says answers to sponsored questions are still generated by Perplexity and are not written or edited by the sponsoring brand.
• Data: the same post says Perplexity "will never share your personal information with advertisers".

Advertiser accounts that can now spend into AI surfaces automatically deserve the same protection as any other high-value login. See the Google Ads Account Security guide for account takeover controls.

What can users see, control or avoid?

User-side commitments are where the providers' statements are most detailed. The table uses the same fields for each product; where a provider has not published a control, the cell says so.

Where a cell reads "not stated", the provider may document the control on a separate privacy or account page; check those pages before telling users that a control does not exist. For workplace guidance on which assistants staff should use and what they should not paste into them, see the Safe Use of AI Chatbots Guide. Organisations that buy enterprise seats can check plan controls in the ChatGPT Enterprise Security Guide and the Perplexity Enterprise Security guide.

How have ads in AI answers developed?

The timeline lists dated official announcements in order. Dates are those printed on the provider's page.

Perplexity's reported end of testing in 2026 is not in the table because it comes from press reporting rather than a Perplexity page.

How to use this tracker

The tracker supports four common jobs. Each uses the same facts in a different way.

1. Plan advertising budgets: start from the markets table, then read the provider's help centre for eligible campaign types. On Google, AI placements come with existing Search, Shopping and Performance Max campaigns, so review negative keywords and brand exclusions before the market goes live rather than expecting a separate AI campaign.
2. Brief legal and brand teams: use the placement and labelling table to show exactly where sponsored content sits and what label users see. Note that Google's AI explainers cannot be edited or approved by the advertiser.
3. Set workplace rules: the user controls table shows which paid plans are ad-free. Where staff use free tiers for work research, an acceptable-use rule should say which plans are approved.
4. Monitor change: record the date of each source you rely on and recheck it before each decision. A market list or tier rule may have changed since the date shown.

How to choose between ad-supported and ad-free AI search

This is a factual decision aid, not a recommendation. The criteria below are the ones the providers' own pages make relevant.

Small organisations often rely on free tiers; the ad and data statements above are the baseline they should read. Large organisations usually buy business or enterprise plans, which OpenAI lists as ad-free, and should still confirm the position for every assistant in use.

Common misreadings

• "Ads change the answer." Every provider on this page states the opposite. The claim to test is labelling and placement, not answer content.
• "AI Mode needs a new campaign." Google's help centre says existing Search, Shopping and Performance Max campaigns are eligible automatically.
• "I can opt out of AI placements on Google." Google states advertisers cannot opt out of AI Overviews or AI Mode separately from Search; negative keywords and brand exclusions remain the levers.
• "Ads are everywhere in English." In-response ads on Google are limited to the listed markets, while ads above or below AI Overviews run wherever AI Overviews exist.
• "All ChatGPT users see ads." OpenAI's announcement limits the test to logged-in US adults on Free and Go, and excludes paid business tiers.

Practitioner checklist

• Record, for each AI assistant your organisation uses, whether its tier carries ads and the date of the source.
• Review negative keywords, brand exclusions and asset pinning on Google campaigns before a new AI Mode market opens.
• Check search terms reports for queries that triggered ads in AI surfaces, remembering the data is aggregated with Search.
• Allow OAI-AdsBot to reach landing pages you submit as ChatGPT ads if you advertise there.
• Tell legal and brand teams that Google AI explainers beside ads cannot be edited or approved by advertisers.
• Name approved, ad-free plans in your AI acceptable-use policy where staff research sensitive topics.
• Protect advertiser accounts with phishing-resistant MFA, since campaigns now spend into AI placements automatically.
• Recheck every market and tier fact against the provider's page before a budget or policy decision.

Standards and references

• Google Ads Help: Ads in AI Overviews and AI Mode
• Google: Ads in AI Mode expand to new markets
• OpenAI: Our approach to advertising and expanding access to ChatGPT
• OpenAI: Overview of OpenAI crawlers
• Microsoft Advertising: Transforming the future of audience engagement
• Microsoft Advertising: An ads ecosystem built for the AI era
• Perplexity: Why we're experimenting with advertising
• Kagi: Pricing

Also checked on 2026-11-15: Search Engine Land: Perplexity stops testing advertising.

Related NHI Mgmt Group resources: AI Answer Engines Hub · Google AI Mode Explained · ChatGPT Search Explained · Microsoft Copilot Web Answers Explained · Perplexity AI Explained