Skip to content
Contact us

Paid traffic

AI-integrated advertising

On this page

AI-integrated advertising means ads that appear in or next to AI-generated answers, for example in ChatGPT and in Google’s AI Mode.

The situation is changing fast. In November 2026, Swedish companies can buy ads in ChatGPT through self-service, while Google’s ads in AI Mode do not yet reach Sweden.

For companies, this means a new place to reach customers while they are comparing options, but also uncertainty about what works.

Why AI advertising is different

Where do the ads appear in ChatGPT and Google AI Mode?

Traditional ads are clearly separated from content

  • Search ads labelled “Sponsored” above the organic results
  • Display ads in sidebars or between content
  • Social media ads labelled “Sponsored” in the feed

ChatGPT shows labelled ads below the answer and keeps them separate from it. Google can show ads in or next to the AI answer, labelled “Sponsored”.

AI ads appear in longer conversations

Unlike search (one question, one answer), interaction with AI often takes place in longer conversations.

Ads can also appear on follow-up questions, which Google has introduced in AI Mode in the US.

Example

User
“What is SEO?”
AI: [Explains SEO]
User
“How long does it take?”
AI: [Explains timelines]
User
“Find me an agency”
AI: [An ad for a specific agency can appear here]

In the example, the ad can appear when the user asks for an agency.

Why ads in AI answers can match the user’s question

Traditional ads are perceived as interruptions.

AI-integrated ads can be relevant because they appear in connection with a specific question.

Requirement: Users must understand what is an ad and what is an organic AI answer. Under the Swedish Marketing Practices Act (marknadsföringslagen), it must be clear that the content is marketing.

Where can you advertise in AI services in November 2026?

Google AI Mode does not yet have ads in Sweden

Google shows ads in AI Mode in the US and in a further ten countries, including Canada and Australia. Sweden is not among them.

How the ads work

  • Text ads and Shopping ads can appear in or next to the answer
  • The ads are labelled “Sponsored”
  • Ads can appear on follow-up questions in AI Mode in the US

For companies: There are no dedicated AI Mode campaigns. Existing Search, Shopping and Performance Max campaigns take part automatically where the ads are available.

ChatGPT (OpenAI) sells ads to Swedish companies

ChatGPT shows ads in the US, in Europe and in several other markets. In August 2026, the ads reached 31 European countries, including Sweden.

Format

  • Labelled ads below the answer, kept separate from it
  • The ads do not affect the answer, according to OpenAI
Status in November 2026
Swedish companies can buy ads through self-service in Ads Manager.
Challenge
The ads only reach users on ChatGPT’s Free and Go tiers. Personalised ads are initially not available within the European Economic Area (EEA).

You cannot advertise on Perplexity

Perplexity has stopped showing ads and has no plans to bring them back.

Not all AI platforms sell ads

Claude (Anthropic) is ad-free, and Anthropic has promised that it will stay that way.

Which ad formats are available in AI services?

How labelled ads appear below the answer in ChatGPT

Example

User: “Best CRM for small businesses?”

Below the answer, an ad labelled as sponsored appears for the fictitious CRM system SampleCRM.

Advantage
The ad is visible in connection with the question.
Challenge
The ad must be clearly labelled so as not to mislead users.

Product ads can appear in AI answers

When users ask about products, Shopping ads can appear in or next to Google’s AI answers.

Example

User: “Running shoes for a marathon”

Next to the answer, an ad labelled as sponsored appears for the fictitious model Runner X, with an image and a price.

Advantage
The user can go straight to purchase.
Challenge
Competition for the same product categories can become intense.

Ads can appear on follow-up questions in Google AI Mode

Advantage: The ad can follow how the user’s needs develop during the conversation.

How targeting works in AI advertising

Targeting works differently in AI conversations than in traditional advertising.

Intent-based targeting starts from what the user asks about

The ad is selected in real time based on the user’s question to the AI.

Example

  • If the user asks about CRM, CRM ads are shown
  • If the user asks about a web agency in Stockholm, local agencies are shown
  • If the user asks about running shoes, sports shops are shown
Advantage
High relevance (the user is actively looking for this right now).
Disadvantage
Competition can become intense for popular topics.

Conversational context can steer the ad

The AI service has access to the whole conversation, not just the latest question.

Example

Earlier in the conversation
User: “I run an online shop on Shopify”
Later
User: “I need help with marketing”
The ad can draw on the context
Below the answer, a labelled ad appears for an agency that works with Shopify.
Advantage
More personal, relevant ads.
Challenge
How much context should the AI use, given privacy concerns?

AI platforms can build user profiles

As in traditional advertising, AI platforms can build profiles based on history.

Data that can be used

  • Previous conversations
  • Interests expressed in AI chats
  • Demographic data (if the user has shared it)

Status: In ChatGPT, personalised ads are initially not available within the EEA, and users can switch personalisation off. The privacy issues are contested.

These are the advantages of AI advertising

You can learn AI advertising early

The channel is new, and advertisers are still learning what works.

AI ads can be highly relevant

Because a conversation provides more context than a single search, ads can be matched to what the user wants to do.

How the matching differs

  • Search advertising: Matches the search term (fairly well)
  • Display: Matches interest or demographics (less precise)
  • AI advertising: Matches the intent in the conversation

AI advertising can reach tech-savvy audiences

AI advertising can be of interest to companies whose audience is

  • Tech-savvy
  • Quick to adopt new technology

For some products and services, that is a valuable audience.

These are the challenges and risks of AI advertising

Formats and prices are uncertain

AI advertising is new. Platforms are still experimenting with formats, pricing and rules.

Risks

  • Formats that work today may disappear tomorrow
  • Prices may shoot up as more advertisers come in
  • The rules on what is allowed may change

How to reduce the risk: Test with small budgets and be prepared to change direction once you have learnt more.

Availability is limited

Availability varies between platforms and countries.

In Sweden, in November 2026, it is possible to buy ads in ChatGPT, while Google’s ads in AI Mode are not yet available here.

Strategy: Start where there is availability and keep an eye on when Google expands to more countries.

Ethics and transparency require clear labelling

It must be clear what is an ad and what is an organic AI answer.

Risks

Recommendation: Always label ads clearly. Put relevance ahead of aggressive selling.

Measurement is less mature

Measurement in AI advertising is less mature than in conventional Google Ads campaigns, although measurement there also has gaps.

Challenges

  • It is difficult to get detailed data
  • Attribution is complicated (how do you link an AI conversation to a later purchase through another channel?)
  • No standardised reporting yet

Strategy: The data is not perfect yet, so track the overall business impact.

How companies get started

Step 1: Monitor the platforms

This is what you should do

  • Keep track of Google Ads updates (AI Mode ads)
  • Follow OpenAI’s news about ads in ChatGPT
  • Read industry news about AI advertising

Goal: Know when formats and rules change and when the ads reach more countries.

Step 2: Define the audience and relevance

Ask yourselves these questions

  • Does the audience use AI assistants?
  • Which questions related to the business does the audience ask?
  • Where in the conversation is an ad relevant?

Example: A company selling enterprise CRM should focus on questions such as “which CRM for 500+ employees” in ChatGPT and AI Mode.

Step 3: Prepare content

Questions in AI services are often longer and more exploratory than ordinary searches.

This is what is needed

  • Clear ad copy: Easy to understand in the context of a longer question
  • Fact-based: Claims that can be substantiated
  • Specific: A clear offer that matches what the user wants to do

Example

Poor AI ad copy
“We are the best at SEO. Contact us.”
Good AI ad copy
“Running an online shop on Shopify? We offer SEO packages from SEK 15,000 per month, including product optimisation and category SEO.”

Step 4: Test with a small budget

Strategy: Start with a minimal budget (SEK 5,000 to 10,000 per month) where there is availability.

Goals

  • Learn how the platform works
  • Test different formats
  • Understand what the audience responds to

Scaling: When something works, you can increase the budget gradually.

Step 5: Measure and optimise

This is what you should measure (where possible)

  • Impressions in AI answers
  • Clicks from AI ads to the website
  • Conversions (purchases or leads)
  • Growth in brand searches (people who saw the ad in an AI service and later search organically)

Optimise

  • Pause underperforming formats
  • Double the budget for what works
  • Test new variants on an ongoing basis

How AI advertising compares with traditional channels

When should you prioritise AI advertising?

Consider AI advertising if

  • The audience is tech-savvy and uses AI tools
  • The business offers something complex that requires advice (an AI conversation suits it better than a banner)
  • There is budget for experiments
  • You want to learn the channel early

Prioritise traditional channels if

  • The audience does not use AI assistants
  • The budget is limited (traditional channels have a proven ROI)
  • An immediate, predictable return is required

AI advertising complements other channels

AI advertising should be seen as a complement, not a replacement.

An example covering the whole buying journey

Awareness

  • Display and social media ads reach a broad audience

Consideration

  • AI advertising reaches people who are actively looking for information in AI tools
  • Search ads catch people who google

Purchase (conversion)

  • Remarketing (display and social media) reminds people who have visited the site but not bought
  • Search ads catch people who are ready to buy

AI advertising belongs in the research phase, when users ask an AI for advice.

How AI advertising may develop

What is likely to happen to AI advertising from 2026 to 2028?

More countries get the ads
Google plans to expand ads in AI Mode to countries including the United Kingdom in the fourth quarter of 2026, and OpenAI has announced more formats and markets.
Standardisation
Platforms will agree on certain formats and ethical guidelines, in the same way that web ads have been standardised.
Increased competition
As more advertisers come in, costs may rise.
Better measurement
Platforms are developing analytics and attribution models, and the data will become more reliable.

What companies can do now

Availability varies between platforms, but the groundwork can be laid right now:

1. Build an organic presence in AI

Optimise content to be cited in AI answers (see AI optimisation).

2. Monitor the platforms

Keep track of when the ads reach more countries and when new formats are added.

3. Experiment with conversational ad copy

Start writing ad creative in a conversational tone. Refine the copy based on what you learn.

4. Budget for tests

Set aside a small part of the ad budget for testing new channels.

Is AI advertising an opportunity or hype?

This is the situation

The ads already exist
ChatGPT and Google show ads in AI services in several markets.
Much is still untested
The platforms are changing formats and terms.

The strategy for most companies

  • Stay informed
  • Test with small budgets when there is availability
  • Scale up what works
  • Still prioritise proven channels (search, display) as the main strategy until AI advertising has matured

For some companies (tech focus, B2B SaaS, AI-related products): AI advertising can deliver a faster ROI because the audience is already there.

For most companies: It is a complementary channel to test, not a replacement for established channels.

Contact us about AI advertising

Carnaby helps you find your way among new ad channels and test AI-integrated advertising where the platforms are available.

Contact us →

Want to know where you stand today?

Call us, or fill in the form and we will get back to you.

Contact us Read about the pre-study