The ethical questions around advertising in AI conversations

Advertising in AI conversations means placing sponsored messages within the answers that AI systems such as Google's AI Overviews or ChatGPT provide to users. This raises immediate ethical questions about transparency, privacy, and the influence marketers exert on people who believe they are receiving a neutral answer.

Key takeaways

• AI conversations create a new advertising channel where the boundary between editorial advice and paid promotion fundamentally blurs.
• Transparency is the central ethical obligation: a clear sponsored label is a minimum requirement, but is structurally harder to anchor in conversational contexts.
• Privacy questions grow as AI systems use personal conversation data for targeting.
• Advertisers have their own responsibility to advertise ethically, regardless of what platforms technically permit.
• AI agencies like AdBrains are already building systems that enable compliant and transparent advertising in AI environments.

What makes advertising in AI conversations different?

In classic Google Ads, the advertisement is clearly separated from organic search results, with a mandatory "Sponsored" label. The user understands the context. In AI conversations, that separation is far less obvious. When an AI system answers a question and weaves in a paid product recommendation, the user initially just sees an answer.

That conversational character makes all the difference. You scan a search result with a critical eye. You trust a conversation partner. Research by the Nielsen Norman Group (2024) confirms that users perceive AI answers as more trustworthy on average than traditional search results. That heightened trust makes the ethical responsibility of advertisers heavier, not lighter.

Traditional ads vs. AI conversation ads

Key characteristics per ad format

Traditional Google Ads

• User clearly sees 'Sponsored' label
• Ad is separate from organic results
• Context: user's search intent
• Fixed ad copy (RSA)
• No conversational embedding
• User freely chooses whether to click

AI conversation adNew ethical question

• Labelling requirements vary by platform
• Ad can be woven into the answer
• Context: personal conversation with shared data
• Dynamically generated recommendation
• Conversational tone increases persuasive effect
• User trusts a 'neutral' answer

Consider a concrete case: someone asks an AI assistant which online course best suits their career switch to the healthcare sector. If HACCP-cursus.com has a paid placement in the AI response at that moment, is the user sufficiently informed about the nature of that recommendation? Only if the platform and the advertiser together ensure clear disclosure, and that is still often where things go wrong today.

The transparency obligation: what is ethically required?

Transparency is the foundation of ethical advertising in AI conversations. At a minimum, a sponsored recommendation must be recognisable as such, regardless of the platform. What this means in practice still varies considerably per environment.

Google has introduced a "Sponsored" label in its AI Overviews product for commercial recommendations, in line with existing Google Ads policies (Google Ads Help, 2026). That is a step in the right direction. But the Dutch Authority for Consumers and Markets (ACM) states in its guidelines for online advertising that sponsorship must always be unmistakably recognisable to the average consumer. Whether a small label in a long AI response meets that standard is an open legal question still actively debated in 2026.

Advertisers themselves also bear responsibility. The following principles help fulfil that responsibility:

• Always require an explicit sponsored label, even if the platform does not enforce it. Discuss this actively with your agency partner.
• Choose relevant contexts: only advertise in AI conversations where your offering genuinely matches the user's question. Irrelevant ads erode trust faster than they generate conversions.
• Avoid manipulative framing: an AI conversation ad must never be worded in a way that makes it appear the AI is making the recommendation from its own conviction.
• Document your campaign settings: in the event of a regulatory inquiry, you want to be able to demonstrate that you consciously chose transparent presentation.

Privacy and data use: where is the line?

Advertising in AI conversations cannot be separated from data collection. To serve a relevant sponsored recommendation, an AI system needs to know what the user is asking, and possibly what they asked before. That makes the privacy question considerably sharper than it ever was in traditional search.

Within the European Union, the GDPR sets clear frameworks for processing personal data for advertising purposes. Consent is in principle required before conversation data can be used as a targeting signal. In practice things get messy quickly: many AI platforms process prompts server-side, leaving users with almost no insight into what actually happens to their input.

What do consumers worry about with AI ads?

Example: distribution of consumer concerns (qualitative overview)

34%

• Lack of transparency34%
• Privacy risks28%
• Manipulative effect22%
• Data use without consent16%

For a company like Clima-Active.nl, which offers air conditioning and heat pump installation and relies on quote requests, the targeting opportunity looks enormously attractive. Someone asking an AI about the best heat pump for a terraced house is an exceptionally qualified lead. That qualification can only be ethically leveraged, though, if the user knows their question feeds into the platform's advertising model and has explicitly consented to that.

The privacy question extends further still, to the use of aggregated conversation data for training targeting models. The AI Act (2024) provides an initial framework for transparency about AI systems, but its specific application to advertising contexts still needs considerable elaboration.

The manipulation question: persuading or influencing?

Every advertisement is designed to persuade. That is ethically acceptable, as long as the user knows they are being persuaded. The line into manipulation is crossed when the advertisement conceals its persuasive character behind the appearance of neutral advice.

In AI conversations, that risk is more real than elsewhere. The tone of an AI system is naturally authoritative and trust-inspiring. If a paid recommendation is embedded within it, the effect on the user's decision may be greater than the same recommendation in a banner or search result. Some researchers describe this as "conversational persuasion," a qualitatively different category from classic advertising influence.

For e-commerce advertisers like ToetsJeKennis.nl, which offers online exams and courses, this is a concrete dilemma. When someone asks which driving exam training is best, and the AI system mentions ToetsJeKennis.nl with a sponsored label, that is transparent and acceptable. When the AI system, however, frames the recommendation as if it were an independent judgment, with the sponsored nature hidden in small print at the bottom, the line toward manipulation becomes very thin.

How AdBrains addresses ethics in AI advertising environments

As an AI-driven Google Ads agency, AdBrains has a distinctive position in this debate. We manage campaigns in environments that are increasingly AI-driven, including Google's Search Generative Experience and AI Overviews. That demands an approach that is technically advanced and ethically sound.

Our multi-agent verification system plays a key role here. Four independent AI agents check every optimisation decision before it is executed. For campaigns that qualify for placement in AI conversation environments, this system automatically verifies whether the ad settings comply with the platform's current transparency requirements. A campaign that would run without a sponsored label in an AI context is blocked before a single impression is served.

Our RSA improvement system actively works on ad copy that is substantively honest. AI analyses Ad Strength scores and automatically rewrites ad texts with POOR status, as well as texts whose framing comes too close to misleading wording. The line "persuasive but not manipulative" is not a vague norm at AdBrains; it is a concrete evaluation criterion in our content review process.

For server-side signal enrichment via our own sGTM infrastructure, we apply strict first-party data principles. We enrich conversion signals exclusively with data for which the end user has given explicit consent through a GDPR-compliant consent flow. This makes our Smart Bidding steering not only more effective, but also sustainable in a regulatory environment that is becoming progressively stricter.

In our practice, we see that advertisers who operate transparently and ethically in AI environments build better Quality Scores over time and generate less wasted spend. The audience that clicks on a clearly labelled recommendation in an AI conversation consistently converts better than traffic attracted through unclear or hidden sponsorship. That is not coincidental. Trust is a conversion factor.

Factors that determine trust in AI ads

Illustrative example: relative importance per factor (scale 0-100)

Clear sponsored label

91

Relevance to the question

78

No personal data used

74

Opt-out option offered

61

Independent review alongside ad

55

For clients like E-4motion.com, which sells new electric folding bikes, our automated search term mining system monitors daily which queries and AI prompts lead to clicks. This allows us to quickly detect whether placements in AI conversation environments contribute to quality traffic or instead generate irrelevant impressions. That data feeds directly into our tCPA/tROAS optimisation, so the budget always lands where the ethical and commercial value is highest.

What this means in practice for advertisers

The ethical questions around AI conversation advertising are not merely philosophical. They have direct practical consequences for how you set up and manage your campaigns.

For lead generation clients like LeroyBrouwer.nl, the above is particularly relevant. The quality of a lead coming in through an AI conversation ad is directly tied to the level of transparency. A user who knows they clicked on a sponsored recommendation and still converts is a significantly more valuable lead than someone who thought they were clicking on neutral advice and was disappointed afterwards.

The road ahead: regulation and self-regulation

In 2026, the regulation of advertising in AI conversations is still very much in development. The European AI Act (2024) sets transparency requirements for AI systems that interact with humans, but the specific application to advertising contexts requires further clarification from regulators. Platforms like Google and Microsoft are also developing their own codes of conduct for advertising in AI environments (Google Ads Help, 2026).

The most sustainable position for advertisers is not to wait for regulation, but to proactively apply standards higher than the minimum requirements. Advertisers who operate ethically in AI conversations build stronger brand reputations over time, attract qualitatively better leads and conversions, and face less exposure in the event of regulatory enforcement.

FAQ: Frequently asked questions about advertising in AI conversations

Is advertising in AI conversations legally permitted?

Yes, advertising in AI conversations is permitted, provided existing advertising law is complied with. Regulators require that sponsorship is always unmistakably recognisable to the average consumer. Platforms like Google are required to display a sponsored label for paid placements in AI Overviews (Google Ads Help, 2026). The AI Act (2024) adds transparency requirements for AI systems that communicate with people, but precise application to advertising embedding is still being worked out by European regulators in 2026.

How does a user know an AI recommendation is sponsored?

In well-configured AI advertising environments, the platform displays a clear "Sponsored" label with every paid recommendation. Google does this for AI Overviews in line with its standard ad labelling policy. In practice, however, the visibility of this label varies by platform and display context. Users should scrutinise the origin of recommendations in AI conversations critically, and advertisers and agencies share responsibility for not hiding or minimising labels.

Can an AI system use my conversation data for ad targeting?

Within the EU, an AI system may only use personal data from conversations for ad targeting if there is a valid legal basis, in most cases the user's explicit consent (GDPR, Article 6). Platforms are required to be transparent about this in their privacy policies. As a user, you can typically manage your consent through the platform's privacy settings. Advertisers must not assume conversation data is freely available as a targeting signal and should verify with their agency partner how data is processed.

What are the risks of not being transparent in AI conversations as an advertiser?

The risks are multiple. Regulators can take enforcement action for non-compliance with transparency requirements, resulting in fines. Reputational damage is at least as harmful: users who feel misled by a hidden sponsored recommendation will quickly develop negative associations with a brand. Moreover, non-transparent advertising yields qualitatively worse leads and conversions over time, as users with wrong expectations drop off after the click. Transparency is therefore not only an ethical obligation, but also a commercially sensible choice.