Ad models behind AI wars: How Google, Meta, OpenAI are approaching advertising

The race to dominate generative AI has largely been framed as a contest over model capability. But for the advertising industry, a different question is beginning to matter just as much: how these new AI interfaces will eventually be monetised.

As platforms including Google, Meta Platforms and OpenAI experiment with generative AI across search, chat and productivity tools, the traditional mechanics of digital advertising are starting to look less certain. AI assistants increasingly answer questions directly rather than sending users to external websites, compressing the discovery journey that digital marketing has relied on for decades.

The stakes are particularly high in markets like India, where performance-driven digital advertising dominates marketing budgets. According to the Pitch Madison Advertising Report 2026, India’s advertising industry crossed roughly ₹1.55 lakh crore in total spend last year, with digital accounting for more than half of that figure. Search and performance-led formats continue to capture a significant share of digital budgets, making any shift in how discovery works online especially consequential for brands.

That shift raises a difficult question for advertisers. If AI becomes the primary interface to the internet, where exactly does advertising fit inside a conversation designed to deliver answers rather than links?

How AI interfaces are the next ad channels

When asked about AI-driven advertising, brand marketers themselves acknowledge that the space is still evolving too quickly for clear strategies to emerge. While most large advertisers are closely watching how generative AI reshapes search, discovery and commerce, many say they are still in an experimental phase, testing AI-led campaign optimisation and creative generation rather than committing significant budgets to new AI-native ad formats. For now, the prevailing sentiment among brands appears to be cautious curiosity rather than immediate large-scale adoption. And also not going on the record.

“The model most likely to scale is the one that integrates ads without breaking the utility of the AI experience,” says Rohit Agarwal, founder and director of Alpha Zegus. He adds that monetisation inside AI environments will likely move closer to intent fulfilment than interruption, as brands attempt to integrate themselves naturally into user intent rather than disrupt it.

Part of the uncertainty stems from the fact that the major AI platforms themselves are pursuing sharply different monetisation strategies, suggesting that AI advertising may evolve through multiple models rather than settling quickly into a single dominant format.

For Google, the strategy is to extend the economics of search advertising into AI-led discovery. The company has already begun integrating ads into AI-generated responses through features such as AI Overviews. According to Google, when contacted by e4m, generative AI is expanding the ways users discover information across search experiences while creating new opportunities for advertisers to appear alongside relevant answers. The company notes that it processes more than five trillion searches annually and that query growth, including commercial queries, continues to rise as AI capabilities improve.

Rather than building an entirely new advertising system for generative AI, Google appears to be layering AI capabilities on top of the world’s existing search advertising infrastructure. Campaign formats such as Performance Max and AI-powered search tools are designed to extend advertiser visibility into emerging AI interfaces without requiring brands to rebuild their strategies from scratch.

Meta Platforms, meanwhile, is taking a different route. Instead of inserting ads directly into conversational AI interfaces, Meta is deploying generative AI across its existing advertising stack to improve targeting, automate campaign optimisation and generate creative assets at scale. The company has been positioning AI primarily as a performance engine for advertisers rather than a new advertising surface. Meta declined to comment for this story.

OpenAI appears to be taking a more cautious approach to monetisation. The company has largely focused on subscription-led revenue through premium ChatGPT tiers while experimenting with integrations and partnerships. OpenAI did not respond to queries sent for this story at the time of publication.

Some AI-native platforms are attempting to differentiate themselves further by stepping away from traditional advertising models altogether. Perplexity AI has experimented with sponsored answers and commerce partnerships within its AI-powered search interface, while Anthropic has largely prioritised enterprise deployments and safety-focused AI development rather than consumer advertising.

The divergence reflects a deeper uncertainty about how generative AI will reshape digital discovery. For much of the past two decades, digital advertising has relied on a relatively simple behavioural chain: users search for information, click through results and eventually land on websites where brands capture attention and conversions.

AI search after the hype 

The stakes are particularly high given the central role digital advertising already plays in India’s marketing ecosystem. According to the dentsu–e4m Digital Advertising Report 2026, India’s digital ad market crossed ₹59,000 crore in 2025, with performance-driven formats such as search, social and e-commerce media accounting for a large share of advertiser budgets.

Search alone continues to remain one of the most reliable high-intent channels for marketers, making any shift in how discovery works online particularly significant. If AI assistants increasingly become the first interface through which users explore products, services and information, the economics of search-led digital advertising could begin to evolve in ways that marketers are only starting to understand.

AI assistants increasingly collapse that journey. Instead of presenting links, they provide direct answers that may satisfy the user’s query without requiring a click.

Dr Siddhant Sethi, an AI specialist and consultant at global advisory firm FSG, says this shift is already beginning to show up in the data. Early research suggests that only around six to eight percent of sessions in Google’s AI Mode result in a visit to an external website. More broadly, studies indicate that roughly 59 percent of Google searches already end without a click.

“These studies measure different products and should not be treated as directly comparable,” Sethi notes, “but they point in the same direction: more discovery and decision-making is happening inside the AI interface rather than on the advertiser’s landing page.”

For advertisers, this shift creates a significant measurement challenge. If users increasingly make decisions inside AI responses, the traditional signals that marketers rely on to evaluate campaign performance may become weaker indicators of influence.

AI search reshaping wellness advertising

Agarwal argues that attribution models will likely need to evolve as a result. For two decades, digital advertising has relied heavily on metrics such as clicks, page visits and session-level tracking to measure campaign performance. But in an AI-first environment, the user may never leave the assistant interface.

“The biggest disruption will be attribution,” Agarwal says. “In an AI-first environment, the traditional funnel breaks because the user may never leave the assistant interface.”

Industry observers believe this could force advertisers to rethink how they measure influence and intent across digital channels. Instead of relying primarily on last-click attribution, brands may increasingly need to track signals such as citation visibility within AI responses, branded search lift and broader indicators of recommendation influence.

There is also a deeper question about whether conventional advertising formats will translate effectively into conversational interfaces.

Vivek Bhargava, co-founder of Consumr.ai, argues that inserting traditional advertising inside generative AI responses may prove difficult to scale. “These systems are designed for conversations and outcomes, not for browsing feeds,” he says, noting that commercial messages inside problem-solving interactions can quickly feel intrusive.

Bhargava believes monetisation models built around task completion may ultimately prove more sustainable. “AI monetisation will succeed only when the recommendation improves the outcome rather than disrupting the conversation,” he says, suggesting that subscription models and agent-driven commerce may become central to the economics of generative AI platforms.

For now, however, the advertising industry is still in the early stages of understanding how generative AI will reshape digital marketing economics. Platforms are experimenting with different monetisation models, advertisers are exploring new measurement frameworks and user behaviour continues to evolve as AI interfaces become more capable.

Are brands building for LLM discovery? 

The race to build the most powerful AI models may dominate headlines today, but the next phase of the battle will likely revolve around monetisation.

For marketers, the challenge is not simply adapting to a new technology. It is understanding how visibility, influence and measurement will work inside systems designed to deliver answers rather than links.

If AI assistants increasingly become the interface to the internet, the advertising industry may need to rethink not just formats but the very mechanics of digital discovery.