Despite once saying he “hated advertising” and that it would be a “last resort”, ChatGPT’s Sam Altman has announced that the AI company is going to start trialling advertising on its free and Go versions of its service in the US in the next few weeks.
While the economics of this are fairly plain – OpenAI projects losses will reach approximately $14bn in 2026, with the company’s cash burn rate remaining around 57% of revenue through 2027, according to financial documents reviewed by Fortune and The Information – the move ushers in a new paradigm for retail media.
With more than 800 million weekly active users, this move transforms ChatGPT into a major competitor to traditional Retail Media Networks (RMNs) by capturing shoppers during the research phase before they arrive on a retailer’s website.
As Google ups the ante in agentic commerce with the roll out of its Universal Commerce Protocol (UCP), the retail world has pivoted towards being increasingly led by AI discovery and away from on-site, banner-heavy advertising to “conversational” ads that follow user intent.
It is easy to suggest that OpenAI needs ads to cover the increasing funding gap as it invests up to $1.4trn in AI infrastructure across the next eight years; it is also clear that brands need AI platforms to carry ads to get their message out there where the consumers are increasingly looking.
The new top of funnel
It is yet to be seen what the ads will look like and how the process will work – the trial isn’t due to start until probably February at the earliest – but from a retail media perspective they have the very real potential of creating a new top of funnel for advertisers. Ads could appear straight after prompts, allowing brands to reach consumers exactly when they are making purchasing decisions, capturing them earlier in the journey.
Ads in ChatGPT could also direct users to specific products, potentially bypassing the sponsored product placements on retailer sites like Amazon or John Lewis.
Some have suggested that ads can also inform responses, with ads appearing not as banners, but as conversational recommendations, placing sponsored products or services directly into the chat response. OpenAI has been very keen to point out that answers remain independent, saying that “Ads do not influence the answers ChatGPT gives you. Answers are optimised based on what’s most helpful to you. Ads are always separate and clearly labelled”.
So, for now at least, the company is trying to keep advertising away from content, but with retail media blurring the lines between advertising and retail and AI blurring the lines between search, discovery and retail, how long can that last? If OpenAI doesn’t do it, someone will.
How it changes ad strategy
With that in mind, how might AI discovery-based retail media advertising change? In simple terms, dramatically. Much as AI discovery changes how websites and SEO work – websites becoming less about SEO and more about being mere repositories of information that AI then draws on – so too does advertising in the AI discovery stream shift what advertisers need to do.
For starters, advertisers must shift from keyword-based bidding to conversational intent maps to ensure that their product data feeds are AI-readable.
Conversational intent maps are visual guides or structures that chart a user’s goals (intents) and the logical paths (flows) to fulfil them in a dialogue with AI, moving beyond keywords to understand the underlying why of a query, allowing for more natural, context-aware, and effective interactions in chatbots, voice assistants, and search. They map user inputs to specific intents, extract key data (entities), and plan the conversation’s progression, handling potential errors and context switches to create smooth, efficient user experiences.
More prosaically, brands and retailers should look to allocate endemic budgets towards off-site content and SEO then that feeds into AI-generated answers, rather than just on-site search ads.
It is also going to be important to think in terms of ‘conversational content’. Unlike static, one-way marketing, such as banner ads or product descriptions, conversational content – often itself powered by Generative AI – is used to engage in real-time, two-way, context-aware conversations that mirror the experience of speaking with, say, an in-store expert. Creative is about to become chatty.
Retailers such as UK electrical retailer Currys, are already using third-party content to attract high-intent shoppers to key category pages, where their own on-site campaigns can close the sale.
Challenges and risks
There is much still to learn of what OpenAI plans to do with advertising and how it will evolve from a retail media perspective, but there are undoubtedly a raft of challenges and risks baked in to such a shift in how advertising works.
The biggest issue, as already stated, is that of trust and bias in AI recommendations. If ads are seen to bias recommendations, user trust in the platform’s objectivity could erode. Ads will likely be embedded within the conversation, such as a product recommendation for a holiday destination when a user asks about travel, but OpenAI insists that they will be in a clearly marked box to avoid interfering with the conversational flow.
But from a retail media point of view, brands are going to want to pay for prominence, which may compromise this principled stance. There is also likely to be a move for brands to want to also pay for prominence in the GPT directory, creating “branded assistants” or, for example, a “Style Assistant GPT”.
While this could be useful for consumers, it does throw up issues around data privacy within such an expanded OpenAI model. OpenAI says that “We [OpenAI] keep your conversations with ChatGPT private from advertisers, and we never sell your data to advertisers. You control how your data is used. You can turn off personalisation and you can clear the data used for ads at any time. We’ll always offer a way to not see ads in ChatGPT, including a paid tier that’s ad-free.”
Interestingly, the company has also pre-empted how advertising may be linked to level of use, saying that “We do not optimise for time spent in ChatGPT. We prioritise user trust and user experience over revenue”.
The road ahead
It is early days with AI/agentic commerce advertising, but the initial consensus appears to be that, while ChatGPT is starting from scratch in the ad world, its massive, active user base makes it a highly valuable, albeit disruptive, new channel for retail media, with early adopters likely to see lower costs of acquisition.
What is for sure is that just as AI discovery is radically changing how we all use the internet – and within that how we shop – it is also now going to radically change how brands need to approach advertising.
That said, OpenAI believes that fundamentally advertising doesn’t change. It says: “The best ads are useful, entertaining, and help people discover new products and services. Given what AI can do, we’re excited to develop new experiences over time that people find more helpful and relevant than any other ads. Conversational interfaces create possibilities for people to go beyond static messages and links. For example, soon you might see an ad and be able to directly ask the questions you need to make a purchase decision.
“Ads also can be transformative for small businesses and emerging brands trying to compete. AI tools level the playing field even further, allowing anyone to create high-quality experiences that help people discover options they might never have found otherwise.”
