Are AI answers becoming ad inventory?

When OpenAI confirmed that it would begin testing advertising inside ChatGPT responses, the announcement marked more than a long-anticipated monetisation move. It signalled a deeper structural shift in how digital advertising may be discovered, evaluated, and valued in an AI-mediated internet.

For nearly two decades, digital advertising has been organised around visibility and interruption. Search ads competed for clicks alongside organic results, while social advertising fought for attention in feeds designed for browsing. Generative AI collapses that logic. In conversational interfaces, discovery, evaluation, and recommendation are compressed into a single answer. Advertising no longer sits around content. It sits inside the decision.

OpenAI’s approach reflects this new reality. Ads will appear at the bottom of relevant responses, clearly labelled and separated from organic answers, and will initially roll out to logged-in adult users on the free and Go tiers in the US. Higher-priced subscriptions will remain ad-free. 

Read e4m report on OpenAI's announcement to test advertising in ChatGPT

Meher Patel, Founder of ad tech platform Hector, frames the shift less as a redistribution of impressions and more as a concentration of influence. “User attention will shift from results scanning to answer acceptance. In traditional search and social, users distribute attention across multiple links and creatives. In AI-mediated experiences, attention concentrates on a single synthesized response and the conversational path that follows,” he says. 

The company has positioned advertising as one of several revenue lines needed to support the extraordinary capital intensity of AI infrastructure, with OpenAI committing to spending roughly $1.4 trillion over the next decade on data centres and computing resources. Advertising revenue is expected to reach the low billions of dollars in 2026, with stronger growth projected thereafter.

Yet the commercial implications stretch far beyond OpenAI’s balance sheet. 

From keyword economics to conversational influence

AI-led search is already disrupting the economics of keyword-driven discovery. As users increasingly rely on AI systems to recommend, advise, and shortlist options, the traditional cost-per-click model is coming under pressure. Fewer clicks are needed when answers are synthesised. Fewer impressions are required when attention concentrates on a single response.

Adtech expert Tejas Tamhane believes this will reshape how marketers allocate budgets rather than trigger a sudden exodus from existing platforms. “Search will be more impacted than social, as AI is increasingly becoming a shopping ally. Users are relying on AI to recommend, advise, and guide their purchase journey. Buying behavior will not become linear though. An AI recommendation does not necessarily equal a purchase. Its real role is inserting products into the consideration set, while final conversion may happen elsewhere,” she says. Until attribution becomes clearer, Tamhane expects budget shifts to remain conservative at 5-10 per cent over the next 12 to 24 months.

Patel estimates that near-term budget movement away from search and social will be incremental, in the range of 0.5-2 per cent of total digital budgets, with higher single-digit shifts possible in high-intent verticals where conversational discovery meaningfully compresses the funnel. 

Measuring impact when clicks disappear

As advertising moves inside AI responses, click-through rates lose their centrality. Influence no longer depends on driving traffic but on shaping decisions within the conversation itself. This forces a fundamental rethink of measurement.

“In a prompt-based advertising environment, click-through rates lose relevance because influence happens inside the response rather than through an explicit action,” says Tamhane. “Success needs to be measured by presence and decision quality instead of clicks. This means tracking how often a brand is mentioned within relevant prompts, its share of voice versus competitors, and whether it appears as a primary or secondary recommendation.”

Patel agrees that accountability will hinge on incrementality rather than interaction. “CTR is a proxy for attention in link-based systems. In conversational environments, the more relevant measure is influence with accountability. Measurement must shift from asking whether the user clicked to whether the AI interaction changed the user’s next action in an attributable and incremental way,” he says. He adds that metrics such as qualified action rates, assisted conversion value, and prompt-to-purchase compression will become central, alongside a visible trust layer that reassures users about ad separation and governance.

For marketers, this reframes effectiveness as consistency of consideration rather than immediacy of response. Brands that appear repeatedly and credibly within relevant conversations may benefit even when conversion happens elsewhere. 

Performance meets brand in a hybrid future

The emergence of AI-embedded advertising also reopens an old debate in a new context. Is this a performance channel, a brand channel, or something in between.

Tamhane argues that the answer depends on category and price point. “AI advertising will evolve into a hybrid model. For online-first brands and small to mid-ticket purchases, it may skew performance-led, as evidenced by ChatGPT’s January 2026 launch of direct checkout with Shopify merchants, where OpenAI takes a 4 per cent transaction fee on completed sales,” she says, referring to Shopify integrations. At the same time, he notes that Google and Microsoft are offering checkout features without additional fees, suggesting the market is still testing which monetisation models will prevail.

Nitin Agarwal, Marketing Head at fintech company Choice Group, sees the balance tilting over time. “In the short term, it will lean toward performance. Marketers will demand accountability, and platforms will need to prove commercial viability. Over time, however, it will become clear that brands with higher credibility and trust are disproportionately favoured by AI systems,” he says. According to Agarwal, early signals already show that users ask AI systems who they should trust rather than who is cheapest, and that overly optimised performance messaging underperforms in conversational environments. 

Trust as the limiting factor

If AI advertising promises high-value moments close to decision-making, trust remains the binding constraint. OpenAI’s insistence on clear labelling, strict separation, and user control over personalisation reflects an awareness that monetisation missteps could undermine long-term engagement.

Patel cautions against overestimating scale too quickly. While AI-embedded advertising could become economically meaningful, a double-digit share of global digital ad spend within five years should be seen as an upside scenario rather than a base case. With global digital ad spend estimated at $678.7 billion in 2025, scaling AI advertising to tens of billions annually would require rapid expansion without eroding credibility.

Agarwal is more bullish on the medium-term opportunity, particularly in high-consideration categories such as finance, education, healthcare, and enterprise software, where AI advice carries disproportionate weight. He believes AI-embedded advertising could reach 10-12 per cent of global digital ad spend within two to three years, driven by value rather than volume.

What is clear is that the old rules of digital advertising are no longer sufficient. As AI systems increasingly mediate choice, advertising is shifting from being seen to being believed. The platforms and brands that succeed will be those that treat monetisation not as an overlay, but as a function of relevance, trust, and decision quality. In that sense, advertising inside AI answers is not just a new format. It is a test of whether the industry can evolve beyond clicks and impressions toward something closer to genuine influence.