ChatGPT Ads: The Next Frontier in Paid Media or a Risk Too Far?

In this article, our Strategist James Doddrell provides some thoughts on how AI advertising will impact our paid media services in 2026. 

On the 9 February, OpenAI announced they are beginning to test ads in ChatGPT in the US. Along with Google introducing ads into AI Mode and AI Overviews in 2025, we at Reading Room are interested in the impact this will have on our client strategies.

Is ChatGPT the next major advertising platform? Should we include LLM ads in our marketing mix? What do AI ads mean for the future of paid advertising?

What Came First, the Search or the Ad?

The search landscape has changed drastically over the last few years. In 2025, 42% of US and European users started their online research with an AI assistant or LLM-based tool rather than a traditional search engine, nearly double the 24% reported in 2024. It’s no surprise now that advertising is now being introduced into LLM search, following in the footsteps of search engines and social media networks.

Advertising has always followed buyer behaviour, not the other way around, and once introduced, it becomes a large stream of revenue generation. From a social media point of view, looking at Meta, revenue generated by advertising in 2025 accounted for 97.62% of all revenue.

So, considering the growth in AI search behaviour, it was inevitable for AI technology providers such as Google and OpenAI to introduce advertising models into their systems. But what does this mean for marketers?

Does My Audience Use AI Search?

If you’re wondering whether your target audiences are using AI search, the answer is, they probably are. A BBC and IPSOS survey in 2025 concluded that one in three use AI weekly. The largest uptake in usage was those aged 55+ and 45-54. So, it’s evident that users across all demographics are embracing AI-powered search as a means of guiding decision-making and discovering and evaluating brands.

With more than a third of users starting searches via AI tools rather than traditional search engines, it’s becoming increasingly important for any organisation to be findable through AI search via content optimisation and site-crawlability. Then when it comes to AI ads, this hypothetically takes that traffic driving potential a step further, serving a sponsored link to your audiences, at exactly the right time. Sounds like a win-win, right? Well, here are our thoughts on the potential risks:

1. Trust Is on Thin Ice

53% of consumers distrust or lack confidence in the reliability and impartiality of AI results. Introducing ads into the mix may only serve to amplify this distrust when users are seeing corporate-sponsored links. Research already shows that trusts collapses quickly when commercial influence is clear in AI outputs.

To mitigate this, OpenAI have tried to dispel concerns, explaining they will always be clearly labelled as sponsored and visually separated from organic answers, as well as ads not displaying near sensitive or regulated topics such as health, mental health or politics. However, is that enough to instil trust in users? It’s for exactly this reason that Perplexity, one of the first gen-AI companies to introduce ads back in 2024, has now abandoned advertising altogether citing fears over eroding user trust.

2. Accuracy Can Be Brought into Question

AI models are known to occasionally hallucinate, generating false, outdated, or misleading information about products, services, or company details. This is already a concern for organic AI-led search, however, add the element of monetised responses to inaccurate organisation information, and you have a recipe for causing irreparable brand erosion.

3. Expect Less Control

AI-led search has led to a loss of brand voice with summaries being neutral and formulaic. But with AI advertising, marketers also seem to have very little control over how ads appear and for what searches.

OpenAI have explained that during the initial roll-out in the US, they will decided which ads to show users. This will be done by matching ads submitted by advertisers with the conversation topic of conversation, chat history, and past ad interactions. Furthermore, if there are multiple advertisers, they will select the one that is most relevant to the user’s chat.

This reduced visibility of how to increase ad performance, as well as potentially high competition for brands competing for limited space in AI summaries may present challenges for brands to optimise for visibility, leading to increased risk of information presenting correctly.

4. Always Be Mindful of Compliance and GDPR

As a full-service digital marketing agency offering website compliance services, compliance is at the front of our mind is, so it’s only natural that we’d look at AI ads from a privacy, data, and GDPR point of view.

LLMs have deep access to user context, historical data, and sensitive information, and as AI chatbots transition into monetisation models that include ads, they are transforming from tools into personalised and persuasive marketing channels. This calls into question how these technologies store and utilise personal data, especially considering GDPR’s core elements regarding data minimisation and purpose limitation.

On the same day that OpenAI started testing ads in ChatGPT, Zoë Hitzig, a researcher at the company, resigned citing concerns over ads being introduced into the LLM. While not calling advertising itself immoral, Hitzig argued that the nature of the data was at stake, with users sharing very personal issues, as well as political and religious beliefs with the chatbot making ChatGPT ads especially risky. Anthropic sought to capitalise on this ethical debate by releasing an anti-ad campaign during the Superbowl in the US.

Furthermore, OpenAI documentation revealed that ad personalisation will be enabled by default for users in the test, which violates GDPR, regardless of whether a user is on a free or paid tier subscription.

Our Technical SEO and Web Compliance specialist, Joel, said:

It doesn’t matter whether you’re on a free or paid tier for GDPR, you can’t legally have it enabled by default when operating in Europe. Consent for ad personalisation must be something users opt into , so it will be interesting to see how OpenAI handle this. Watch this space!

When rolled out to Europe and other regions with strict data protection regulations, advertising in platforms like ChatGPT may not automatically make a brand GDPR compliant. This could introduces additional risks and responsibilities which brands will have to manage proactively.

ChatGP Ads In, Paid Search Out?

So, what does this mean for our paid strategies? Should you be running LLM ads as soon as possible? What does it mean for other paid channels? These are all questions that we are considering every day as this technology rolls out, and we have finished this article with our thoughts below.

1. If Trust Is Maintained, AI Ads Are a Valuable Consideration for Your Channel Mix

The 2025 Consumer Adoption of AI report found that nearly 41% of consumers trust Gen AI search results more than paid search results. As paid strategists, this shift in consumer trust and how trust will be impacted by the introduction of more advertising within LLMs is very interesting to observe. If consumer trust builds for LLM ads, this could have a large impact on organisations that rely on paid search as a major part of their customer acquisition and sales strategy

2. Don’t Replace One with Another

With ChatGPT Ads being matched to relevant conversations, this suggests formats will include contextual recommendations, paid follow-up prompts, or affiliate or referral links. In practice, this means advertising in LLMs will feel less like paid search, and more like paid influence over recommendations. Therefore, instead of replacing paid search, LLM ads potentially provide an additional opportunity to build awareness and influence during your audience’s consideration phase.

3. It’s Worth a Test as Long as You Understand the Risks

On the one hand, ChatGPT and other LLMs present a high-intent environment where highly engaged users are seeking solutions via conversational and descriptive prompts. On the other, we have highlighted some considerable risks, and the technology is very much in its infancy meaning reach and ROI may be limited, with unpredictable costs per result, and potentially limited measurement. As long as your experts feel informed, and businesses are not expecting a 200% ROI from the outset, AI Ads provide a valuable chance for marketing experimentation and innovation.

Our Final Thoughts

The topic of advertising in LLMs is one that’s constantly evolving, with organisations on one side launching tests to bring ads to chatbots, while others abandon advertising strategies all together, all amidst multiple notable departures of AI experts from prominent tech-players (including Anthropic and Elon Musk's xAI).

For Reading Room, audience safety is paramount to maintain brand respect and trust for our clients and therefore, we will continue to monitor developments in this space during this period of commercialisation to fully understand the facts, before encouraging our clients to start advertising within LLMs.