Generative‑AI systems have moved from experimental tools to the way many people search, shop and get recommendations. Google’s AI Overviews and AI Mode, Microsoft’s Copilot, OpenAI’s ChatGPT, Perplexity, Meta AI and subscription‑based search engines like Kagi and You.com all offer conversational responses rather than lists of links. As these tools scale, the big question for marketers is how and when advertising will be integrated, and whether the new formats will help or hurt businesses that rely on search traffic.
This article reviews the current state of ads in AI search and chat platforms as of November 18 2025. It summarizes the roll‑out timeline for each platform and explores the benefits and disadvantages of placing ads in generative AI experiences.
Google: AI Overviews & AI Mode
Roll‑out timeline
| Event | Date | Details |
|---|---|---|
| Initial experiments in AI Overviews (United States) | May 2024 – Oct 2024 | Google quietly began testing ads displayed above and below AI Overview summaries in the U.S., particularly on mobile. The company said ads would help users go from discovery to decision faster and that they were only shown when relevant. Philipp Schindler (Google) later announced that ads would roll out to other platforms and countries. |
| Google Marketing Live 2025 | May 21 2025 | At GML 2025, Google officially announced ads inside AI Overviews. VP Vidhya Srinivasan said the Gemini model could detect subtle commercial intent and that connecting a user to a product through an ad could be the “most helpful next step”. Google also previewed AI Mode, a conversational interface with follow‑up questions where ads may appear below or integrated into responses. |
| Expansion to desktop and new markets | June 2025 – Sept 2025 | Google expanded the pilot to desktop and began experimenting with ads embedded within AI Mode responses in the U.S.. By late 2025 the company reported that ads in AI Overviews were live in English in the U.S. and would soon expand to other English‑speaking countries. |
| Google Access/October 2025 update | Oct 13 2025 | At a Google Access event, the company confirmed it would roll out ads in AI Overviews beyond the U.S. to select English‑speaking markets by the end of 2025. Google said the gradual rollout would give advertisers time to adapt. |
How Google’s ads work
Placement – Ads can appear above, below or within the AI Overview. Ads within the summary use both the user’s query and the AI‑generated content to determine eligibility. Google uses existing search ad auction signals (Quality Score, Ad Rank, etc.) for above/below placements and adds generative content relevance for in‑summary ads.
Eligibility – Only campaigns using Google’s AI‑driven products (Performance Max, AI Max, broad‑match Search and Shopping) can show in AI responses. Advertisers cannot opt out or bid specifically for AI Mode.
User experience – Google says ads shorten the path from discovery to purchase and are clearly labeled to avoid confusing organic and sponsored content. Ads within AI Mode responses appear as part of the conversational answer, guiding users to products relevant to the discussion.
Benefits
Quicker path to conversion – Google positions ads in AI responses as the “next best step.” When the AI explains how to clean a green pool, the ad may offer a pool vacuum. This contextual placement can lead to fewer clicks and faster conversions.
High intent and relevance – Ads inside AI responses leverage the user’s question and the AI’s interpretation, potentially increasing relevance and engagement. Early experiments show that context‑matched ads provide more qualified traffic and could mitigate the click‑through rate (CTR) losses seen when AI Overviews displaced traditional results.
New real estate for advertisers – With fewer traditional “blue links,” AI responses open new surfaces for paid discovery. Brands that embrace broad match, Performance Max and high‑quality data feeds can appear in previously unreachable conversational contexts.
Disadvantages & controversies
Reduced organic traffic – Studies have reported a 34.5 % drop in clicks when AI Overviews are present. While ads may recover some lost visibility, many publishers fear decreased organic traffic to their sites.
Opaque reporting and control – Google currently does not provide separate reporting for ad performance inside AI Mode or AI Overviews. Advertisers cannot target AI Mode specifically, making optimization difficult.
Dependence on broad match & automation – To show in AI results, advertisers must rely on Google’s automated campaigns. Marketers have expressed concern that this reduces control and may increase spend without clear attribution.
User trust and accuracy – Blending ads with AI answers raises concerns that commercial influences might bias responses. Google states that generative summaries are not influenced by ads, but critics argue that the line between helpful recommendation and advertisement could blur.
Microsoft: Copilot & Conversational Ads
Roll‑out timeline
Microsoft began building its AI companion Copilot into Bing search, Edge and Windows in 2023. Unlike Google, Microsoft positions Copilot as a conversational search assistant integrated into its advertising ecosystem.
| Event | Date | Details |
|---|---|---|
| Microsoft Advertising Accelerate 2024 | March 5 2025 | Microsoft announced core principles for ads in Copilot. It promised a user‑centric approach where organic responses are never influenced by sponsored content and ads are distinctly marked. It also introduced Showroom ads and Dynamic filters – interactive ad formats built for conversation. |
| Ramp‑up across markets | March 2025 | Microsoft disclosed that ads in Copilot have fully ramped in all English-, French- and German‑speaking markets, with Spanish and Japanese launching in April 2025 and more languages to follow. Ads leverage entire conversations to improve relevance, delivering ad relevance metrics 25% higher than traditional search. |
| Microsoft Advertising research (Aug 2025) | Aug 6 2025 | Microsoft reported that Copilot’s intuitive ad formats drive 73% higher click‑through rates and 16 % higher conversion rates compared with standard search ads. Customer journeys were 33 % shorter, indicating fewer steps to conversion. |
How Copilot ads work
User‑centric principles – Microsoft states that ads appear only when relevant and do not influence the AI’s organic answer. Sponsored content is clearly labeled, and queries remain private under Microsoft’s data policies.
Showroom ads & Dynamic filters – Showroom ads create an immersive space where users can explore a product through interactive content and, in the future, speak with a brand agent. Dynamic filters allow users to refine a product without typing additional questions.
Dynamic ad generation – Microsoft uses query signals and conversation context to dynamically generate ad copy and images. Early tests improved click‑through rates by 20 %.
Benefits
Higher engagement and conversion – Microsoft’s research shows that conversational ads deliver significantly higher engagement and conversion metrics compared with traditional search ads. This suggests that users respond positively to context‑aware recommendations.
Immersive shopping experiences – Showroom ads provide a digital space akin to a store visit, where users can examine products and ask follow‑up questions. Future iterations will allow interactions with brand agents, offering a blend of commerce and customer service.
Global availability – By mid‑2025 Microsoft had ramped ads in multiple languages, providing advertisers with broad reach.
Disadvantages & concerns
Reliance on Microsoft’s ecosystem – To appear in Copilot, advertisers must use the Microsoft Advertising platform and adopt dynamic ad generation, which may be challenging for small advertisers.
Intrusiveness vs. convenience – While Microsoft claims that ads do not influence the AI’s responses, some users may find interactive formats like Showroom ads intrusive or distracting.
Data privacy – Ads rely on conversation data to tailor recommendations. Although Microsoft promises privacy, increased data collection raises concerns about profiling and surveillance.
Perplexity AI
Roll‑out timeline
| Event | Date | Details |
|---|---|---|
| Initial ads launch | Nov 12 2024 | Perplexity, an AI‑powered answer engine, started experimenting with sponsored follow‑up questions and paid media beside answers. The company stressed that ads would be clearly marked, would not influence the content of answers, and that no personal data would be shared with advertisers. Early brand partners included Indeed, Whole Foods Market and Universal McCann. |
| Aggressive expansion | Early 2025 | Perplexity expanded its ads program to more U.S. users, seeking to diversify revenue beyond subscriptions. Ads remained limited to the U.S. and generated modest revenue (≈$20,000 out of $34 million). |
| Pause on new ads | Oct 2025 | Facing user backlash and low revenue, Perplexity halted onboarding new advertisers to “avoid inundating users with ads” and rethink how advertising fits into its product. Jessica Chan, Perplexity’s head of marketing, said the company would scale thoughtfully and rebuild trust and measurement metrics. |
Benefits
Sustainable business model – Perplexity’s CEO noted that subscription revenue alone would not sustain the platform, and ads were necessary to keep the service free.
Contextual follow‑up – The ads appear as follow‑up questions tailored to the user’s inquiry, which can feel more natural than display ads.
Clear labeling – Perplexity emphasizes that sponsored results are labeled and that the engine’s answers are not influenced by advertisers.
Disadvantages
User experience – Some users complained that the ads disrupted the minimal interface. Perplexity acknowledged this by pausing new advertisers and rethinking its approach.
Limited revenue – Despite the experiments, ads generated little revenue, suggesting that the model may not scale without harming user trust.
Measurement challenges – Advertisers struggled to measure performance within the conversational environment, leading to hesitancy to invest further.
OpenAI’s ChatGPT
Current status & statements
OpenAI has not yet launched advertising within ChatGPT, but internal planning documents and CEO comments indicate a future direction:
Internal projections – Reporting in early 2025 indicated that OpenAI expected to earn about $1 billion from free‑user monetization (ads) in 2026 and potentially $25 billion by 2029. These projections suggested that OpenAI is “staffing up to turn ChatGPT into an ad platform”.
Sam Altman’s evolving views – In 2024 Altman called ads a “last resort.” By mid‑2025 he said he likes Instagram ads and could envision a “cool ad product” that adds value to users. Altman emphasised that he would only accept ads that do not compromise trust and hinted at affiliate‑fee models where ChatGPT recommends a product and takes a small commission if the user books.
November 2025 interview – In an interview on Conversations with Tyler, Altman confirmed that ChatGPT will likely try ads “at some point,” but he still has no idea what they will look like. He argued that Google’s model profits when search fails and vowed that ChatGPT would avoid pay‑to‑play answers. Altman suggested that ChatGPT might present the best recommendation first and take a commission only if the user purchases, rather than accepting payment to downgrade results.
Implications
OpenAI’s approach appears more cautious than Google’s. Ads are unlikely to appear within ChatGPT before 2026, and any implementation may be tied to commerce actions rather than display ads. Benefits of such a model could include maintaining trust by only recommending the best options and earning revenue through commissions. Disadvantages include the potential for hidden biases if commercial relationships influence recommendations and the challenge of balancing user privacy with monetization.
Meta AI: Personalized Ads From Conversations
Roll‑out timeline
| Event | Date | Details |
|---|---|---|
| Policy notification begins | Oct 7 2025 | Meta started notifying users that their interactions with Meta AI across Facebook, Instagram, Messenger, WhatsApp and Ray‑Ban smart glasses would soon be used to personalize content and ads. |
| Policy takes effect | Dec 16 2025 | Meta will begin using users’ AI chat interactions (text or voice) as targeting signals for ads and content recommendations. Users cannot opt out; the change applies globally except in the UK, EU and South Korea. Interactions will be combined with existing signals (likes and follows) to shape ads, but conversations about sensitive topics—religious views, sexual orientation, political views, health conditions, racial/ethnic origin or philosophical beliefs—will not be used. |
Benefits
Richer intent signals – Meta’s marketing materials note that AI chats reveal explicit intent (“Which hiking boots are good for beginners?”), enabling more precise targeting than passive signals like page likes. For advertisers, this could mean reaching users at the exact moment they express interest.
Unified ecosystem – Meta’s vast social graph allows AI‑derived signals to flow into ads across Facebook, Instagram, Messenger and WhatsApp. This integration could improve ad performance and keep Meta competitive with Google and Amazon.
Massive reach – Meta AI already has about one billion monthly active users. Leveraging conversations could significantly boost the value of its ad inventory.
Disadvantages & concerns
Privacy backlash – Users cannot opt out unless they stop using Meta AI. Academics and privacy advocates warn that using seemingly private conversations for ad targeting constitutes “surveillance under the guise of personalization”. Users might avoid or manipulate the AI to prevent data collection, reducing targeting quality.
Sensitive data – Although Meta excludes certain categories, conversation data could still reveal personal circumstances. The policy’s global roll‑out (excluding only a few regions) heightens concerns about data protection laws.
Uncertain ad placement – Meta has not announced ads inside the AI chat itself; instead the conversation data will inform ad targeting across feeds and Stories. If Meta later serves ads directly in chat, user trust could erode further.
Platforms that avoid ads: Kagi & You.com
Kagi
Kagi is a premium search engine launched in 2022. Its business model is subscription‑based: no ads, no tracking and no sponsored results. The website states that Kagi “doesn’t sell your attention to advertisers” and “doesn’t clutter your results with sponsored content”. Kagi positions itself as user‑centric search and charges $5–$25 per month, eliminating the incentive conflicts present in ad‑supported models. This model appeals to privacy‑conscious users and those who distrust algorithmic ranking influenced by advertising.
You.com
You.com similarly positions itself as an ad‑free AI search engine. Its founders emphasize a privacy‑first, user‑controlled experience where users can pin or block sites and the company generates revenue through premium subscriptions and API access rather than advertising. By avoiding ads, You.com claims to deliver unbiased answers and protect user data. The downside is that its growth depends on subscription revenue, which may limit scale.
Summary: When will ads appear and what does it mean for marketers?
| Platform | Current status (Nov 2025) | Expected future | Benefits | Disadvantages |
|---|---|---|---|---|
| Google AI Overviews & AI Mode | Ads appear above/below AI Overviews and are being tested within summaries in the U.S.; ads inside AI Mode (conversational chat) are being tested in the U.S., with international roll‑out planned for late 2025. | Google plans to expand to select English‑speaking markets by the end of 2025 and likely to other countries in 2026. | Contextual ads shorten the path from question to purchase; high intent; new inventory for advertisers; potential to recover lost CTR. | Loss of organic traffic; opaque reporting; mandatory use of automated campaigns; risk of eroding trust if ads bias answers. |
| Microsoft Copilot (Bing/Windows) | Ads fully rolled out in English, French and German markets, expanding to Spanish and Japanese; interactive formats (Showroom ads, Dynamic filters) piloting. | Ads will extend to more languages and deeper AI‑generated creative tools. Microsoft expects higher CTR and conversion rates to continue. | Higher engagement and conversion; immersive shopping experiences; clear labeling; global reach. | Requires reliance on Microsoft’s platform; interactive ads may feel intrusive; conversation data used to target ads raises privacy concerns. |
| Perplexity AI | Sponsored follow‑up questions and side‑of‑answer ads introduced in Nov 2024; expansion in early 2025; pause on new advertisers in Oct 2025 to rethink strategy. | Perplexity is reassessing its ad model; ads may return in a scaled‑back form or shift to subscription-only revenue. | Provides revenue to support free service; ads are clearly marked; contextual follow‑ups feel natural. | Generated little revenue; user experience concerns; measurement issues; currently paused. |
| OpenAI ChatGPT | No ads in ChatGPT yet. Internal documents forecast $1 billion from ads in 2026. Sam Altman says ChatGPT will test ads “at some point,” but details are unknown. | Ads may debut in 2026 with a likely affiliate‑fee model where ChatGPT recommends products and takes a small commission. | Could preserve trust by recommending the best product first and charging only on purchase; avoids cluttering conversation with banner ads. | Risk of hidden commercial influences; requires balancing revenue with unbiased recommendations. |
| Meta AI | Starting Dec 16 2025, Meta will use AI chat interactions to personalize ads across Facebook, Instagram, Messenger and WhatsApp; no opt‑out. | Meta may eventually show ads within the chat interface, but currently it only uses chat signals to inform ads elsewhere. | Rich intent data gives advertisers explicit signals; unified cross‑platform targeting; enormous reach. | Significant privacy concerns; potential for user backlash; sensitive topics excluded but still invasive; no opt‑out. |
| Kagi & You.com | Ad‑free search engines. Kagi’s subscription model means no ads and no tracking. You.com similarly avoids ads and monetizes through premium subscriptions and API access. | Likely to remain ad‑free, relying on subscription growth. | No incentive conflicts; privacy‑focused; unbiased results. | Smaller user base; limited resources; subscription fees may deter users. |
Key takeaways for marketers
Prepare for conversational advertising – Conversational AI is rapidly becoming the default search experience. Marketers should test Performance Max, AI Max and broad match campaigns to increase eligibility for Google’s AI placements, and explore Microsoft’s Copilot ads.
Focus on context and intent – Ads within AI responses use richer signals than keyword queries. Successful campaigns will align product data and creative assets with the questions users ask, rather than relying solely on keyword bidding.
Invest in high‑quality data feeds – AI systems prefer ads with accurate product information and diverse creative assets. Maintaining clean feeds and up‑to‑date assets will improve relevance scores.
Monitor privacy and transparency – Platforms like Meta face backlash for using chat data without opt‑out. Advertisers should stay informed about privacy policies and be prepared to adjust targeting if consumers rebel against surveillance.
Balance reach and reputation – While new ad formats offer reach and efficiency, they also risk eroding trust. Brands should ensure their offers genuinely help users; otherwise, intrusive ads could harm reputation in an era when users can switch to ad‑free alternatives like Kagi and You.com.
FAQ
When will ads start appearing in AI chat platforms and search experiences?
Most platforms are experimenting with ads now. Google began testing ads in AI Overviews in 2024 and plans to expand them internationally by late 2025. Microsoft’s Copilot advertising is already live in several languages. Perplexity introduced ads in late 2024 but has paused new advertisers. ChatGPT has not yet rolled out ads but may do so around 2026. Meta will start using AI chat interactions for ad targeting on December 16 2025.
How are ads in AI Overviews and AI Mode different from traditional search ads?
Unlike traditional search ads that appear as separate listings, ads in AI Overviews and AI Mode can be embedded within the AI‑generated answer or displayed immediately above or below it. The placement is determined by the AI’s understanding of the user’s question and the relevance of the advertiser’s content. This makes the ads feel more like recommendations and less like distinct advertisements.
What are the benefits of placing ads in AI search and chat experiences?
The main advantage is relevance: ads can respond directly to a user’s query, shortening the path from question to purchase. Conversational ads can have higher click‑through and conversion rates because they leverage context and intent. They also create new advertising real estate at a time when traditional search results pages are shrinking.
What are the downsides of ads in AI platforms?
Blurring ads with organic AI responses risks eroding trust if users cannot distinguish between recommendations and paid placements. Advertisers also lose some control over targeting because campaigns must be automated to qualify for AI placements. There are additional concerns about user privacy, especially when chat interactions are used for ad targeting without an opt‑out.
Are there any AI search engines that don’t use ads?
Yes. Subscription‑based services like Kagi and You.com do not show ads at all. They generate revenue from subscriptions and API access, offering an ad‑free, privacy‑focused experience.