AI isn’t just changing search — it’s rewriting the ad stack

While there has been fear about the impending impact to web traffic, not enough people are talking about the significant opportunity this trend creates for the advertising ecosystem. The question isn’t whether these platforms will monetize; it’s whether they’ll build the next generation of walled gardens or integrate with the open web infrastructure that has powered digital advertising for two decades.

Based on ongoing conversations with both established and emerging generative AI platforms, the monetization decisions being made right now will determine the structure of digital advertising for the next decade. And the window for influence is closing fast.

The monetization maturity curve

Generative AI platforms are at different stages of monetization sophistication. Anthropic’s Claude focuses on enterprise API access. OpenAI has introduced commerce integrations, such as Shopify partnerships, that enable product recommendations within ChatGPT. Perplexity’s advertising efforts remain in flux following early pilots and leadership transitions, with a minimal current revenue impact. Yet, continued funding and audience growth create a foundation for renewed exploration of monetization.

These approaches reveal a consistent pattern: Every platform is converging toward hybrid models combining subscriptions with advertising. As Stratechery’s Ben Thompson has argued, scaled consumer platforms ultimately need ad-supported models alongside subscriptions to reach a global audience. The computing costs alone demand multiple revenue streams.

The more interesting question is which infrastructure will power it.

Why generative AI advertising will be different from search

The default assumption is that generative AI advertising will follow the search playbook: high-intent queries matched with sponsored results and cost-per-click economics.

But generative AI interactions are fundamentally different. When someone Googles “best hiking boots,” they receive 10 blue links. When they ask ChatGPT the same question, the AI provides a definitive answer with specific product recommendations. The platform isn’t facilitating choice; it’s making the choice.

This changes everything. Generative AI advertising will likely resemble retail media: contextual, commerce-driven and optimized for conversions rather than clicks. The challenge is execution. This requires real-time semantic targeting, AI-native brand safety frameworks and direct access to premium demand in high-intent categories like e-commerce and travel, supported by optimization models that understand and price against true user intent within conversational queries.

Generative AI platforms will require thoughtful approaches to monetization

Generative AI platforms are evaluating whether to build proprietary ad systems or partner with existing infrastructure. The build option is tempting: higher margins and complete control.

But there’s a hidden risk: advertiser fragmentation.

If every platform builds closed ad systems, advertisers face dozens of incompatible platforms requiring custom integrations and separate measurement frameworks. Many will simply deprioritize these channels.

Platforms that integrate with established programmatic infrastructure, while maintaining control over user experience and data, gain immediate access to scaled demand. These benefit from competitive pressure as advertisers optimize budgets toward top performers. And they avoid building complex ad tech from scratch.

The challenge is finding partners with the right infrastructure. Not every sell-side platform is equipped for generative AI monetization.

Success requires three specific capabilities. First, native format expertise. Generative AI advertising cannot look like banner ads. It requires formats designed for conversational interfaces, including in-feed placements, contextual product recommendations and sponsored responses that feel native to the chat experience. Supply partners with deep experience in retail media, mobile native, in-feed and CTV environments understand how to balance monetization with user experience and how to navigate complex product catalogs to deliver relevant, high‑quality recommendations.

Second, commerce and retail media infrastructure. The highest-value generative AI queries are commerce-driven. A user searching “best hiking boots under $200” represents peak purchase intent. Monetizing these moments requires direct relationships with retail advertisers, the ability to tie ad exposure to downstream sales and targeting that incorporates commerce signals. Partners that already power retail media for major e-commerce platforms bring this capability ready-made.

Third, curated marketplace models. Generative AI platforms cannot afford to compromise user trust with low-quality ads. Rather than open exchanges where any advertiser can bid, these platforms need curated marketplaces with pre-vetted premium demand and strict quality controls. This approach, already proven across premium publishing and CTV, ensures ads enhance rather than detract from the user experience.

The big infrastructure question

The advertising industry is facing its most consequential fork since the rise of programmatic trading. Generative AI platforms will choose their monetization partners over the next two years. Those partnerships will determine whether the next era of digital advertising is open and ecosystem-inclusive — or replicates the walled garden dynamics that have frustrated the industry for a decade.

The platforms that understand this aren’t just choosing ad vendors. They’re choosing which version of the internet they want to build.

The infrastructure decisions made in 2025 will shape digital advertising for the next decade. The window to influence those decisions won’t stay open for long.

Partner insights from PubMatic