AI in Advertising: The gap between hype and reality

Walk into any agency pitch today and you will almost certainly hear the phrase ‘AI-first’. It sits in the deck somewhere between the credentials slide and the strategy reveal, usually accompanied by a flowchart of tools, a few logos from OpenAI or Google, and the implicit promise that this agency, above all others, has cracked the code on artificial intelligence. It sounds impressive. It is meant to. The question that rarely gets asked in the room (and that the industry is only now beginning to reckon with seriously) is whether any of it is true.

The gap between how AI is being sold and how it is actually being used inside agencies is wider than most would publicly admit. According to the 2024 WFA Global Marketing Survey, over 70% of marketers globally said they were actively exploring or deploying AI tools, yet fewer than a third reported meaningful integration into core creative workflows. In India, where the advertising industry is projected to touch ₹1.64 lakh crore in 2025 according to the Pitch Madison Advertising Report, the rush to appear technologically sophisticated has intensified. But sophistication in a pitch deck is not the same as sophistication in delivery.

The pitch deck problem

Shradha Agarwal, Co-founder and Global CEO of Grapes Worldwide is candid about where the industry currently stands, saying, “There is definitely a gap between what is showcased in pitches and what is consistently embedded into day-to-day workflows. AI today is being actively used across specific parts of the creative process such as research, content adaptation and performance optimisation. End-to-end integration is still a work in progress.”

That distinction between selective deployment and systemic integration is crucial. Most agencies using AI well are doing so in discrete pockets: a copywriter using a generative tool to produce ten variants of a headline, a data team running AI-assisted audience segmentation, a content team using machine translation to adapt creatives for regional markets. These are real uses. They deliver real efficiency. But they are a long way from the ‘AI-first’ promise, which implies that intelligence is the engine, not just a component.

Agarwal acknowledges the industry’s tendency to overstate its hand. “There’s been some tendency to overplay the ‘AI-first’ narrative,” she says, “but in reality, adoption is moving step by step. The focus today is on using these tools where they genuinely make a difference, rather than positioning them as a one-size-fits-all solution.” The honesty is refreshing. The problem is that honesty of this kind is rarely what makes it into a pitch.

AI everywhere, less to stand out

The AI noise is not limited to agencies. Brands, too, have caught the affliction, and increasingly, the people marketing AI-powered products are discovering that the language has become so overused it has lost all meaning. Raj Sarkar, Chief Marketing Officer at CloudBees, a leading software delivery platform, puts it with the bluntness of someone who has watched the cycle play out in real time. “Every company I talk to now has ‘AI’ in its messaging — not as a differentiator, but as a default. Suddenly, everything is ‘AI-powered,’ ‘AI-native,’ or ‘AI-first.’ And in the process, every brand sounds exactly the same.”

This is what happens at the inflection point of any major technology cycle. A genuine capability emerges, early adopters gain real advantage, and then everyone else scrambles to claim the same territory, regardless of whether they have actually built it. The signal degrades into noise. In advertising, where differentiation is the entire point, the irony of an industry homogenising its message around a ‘differentiation tool’ is not lost on sharper observers.

“This is what happens when marketers confuse technology with story, and speeds and feeds with value,” says Sarkar. “We’ve traded narrative for noise, and relevance for buzzwords.” His prescription is straightforward: “Customers don’t buy what the technology is — they buy what it does for them. When everyone shouts ‘AI,’ no one stands out. The best marketers don’t just talk about AI, they talk about what it does for customers.” It is advice that applies as much to agencies pitching their AI credentials as it does to brands broadcasting their AI-powered products.

Clients are asking sharper questions now

The good news (if there is good news here) is that the market is self-correcting, and clients are driving the correction. The early phase of AI in pitches was marked by curiosity on both sides. Agencies brought it up because it sounded progressive; clients listened because they were unsure what to ask. That phase is over. A new, more demanding conversation has begun.

Agarwal notes the shift clearly. “Clients today are far more aware of AI and are asking sharper questions around efficiency, speed and measurable outcomes. Agencies did push this conversation early on by bringing it up more often in pitches. That early curiosity is settling. Clients are now getting into specifics around timelines, budgets and output. The focus has clearly moved to what actually works in everyday execution.”

This is a significant development. When clients start asking ‘show me the workflow’ instead of ‘tell me about your AI strategy,’ agencies can no longer survive on narrative alone. The accountability loop is tightening, and that is ultimately healthy, not just for clients, but for agencies that have actually done the work of building real capability. It separates the ones who were always doing the work from the ones who were only selling the story.

India's edge: Implementation over invention

There is a parallel conversation happening at a macro level that has direct implications for where Indian agencies and tech firms fit in the global AI story. The narrative has largely been shaped by the West: Silicon Valley building foundation models, European regulators debating guardrails, and everyone else trying to keep up. But practitioners on the ground in India are beginning to articulate a different kind of advantage.

Jaimin Prajapati, CEO of WebBrains Technologies, a product and AI engineering firm that builds custom AI solutions for businesses across retail, logistics, and healthcare, frames it as a question of where the real value gets created. “The biggest challenge isn’t the AI model. It’s applying AI inside real businesses,” he says. His view, shaped by conversations with founders and engineers at the 2026 AI Impact Summit, is that India is not trying to build the next OpenAI. Something different is happening. “India is becoming the AI implementation capital of the world.”

The distinction matters enormously for the advertising industry. Foundation model development requires billions in capital and deep compute infrastructure. Implementation requires engineering talent, contextual knowledge, and the patience to make messy real-world data actually work. India has an outsized claim on all three. “That gap between AI hype and AI reality is where most of the value will be created. And honestly, that’s where India is strongest,” says Prajapati.

For Indian agencies, this reframe is useful. The pressure to appear ‘AI-first’ in the Western sense (building proprietary models, acquiring AI startups, and launching named platforms) may be misplaced energy. The more durable competitive advantage may lie in becoming genuinely excellent at deploying AI into client businesses: personalising at scale, automating content localisation across India’s extraordinary linguistic diversity, and turning performance data into creative decisions faster than any manual process could.

The road from differentiator to default

The trajectory of AI moving from pitch differentiator to default part of the creative process that Agarwal describes is where the industry is ultimately headed. The question is how long the transition takes and how much credibility gets spent in the meantime on claims that cannot be substantiated.

The Anthropic Economic Index released in early 2025 found that AI is currently augmenting rather than replacing creative roles, with the heaviest usage concentrated in research, drafting, and iteration tasks. That pattern matches what practitioners on the ground are reporting: AI as a tireless junior collaborator, not as a creative director. The implication is that the agencies getting the most value from AI are those that have redesigned their workflows around it, not those that have simply added it to their pitch vocabulary.

Agarwal offers what might be the most grounded outlook available. “Over time, as tools become more reliable and teams more fluent, this gap will narrow and AI will move from being a pitch differentiator to a default part of the creative process.” That is almost certainly correct. The agencies that will be well-positioned when that moment arrives are not necessarily the ones making the loudest claims today. They are the ones quietly redesigning how their teams work, one workflow at a time.

In advertising, the story has always mattered. But when the story gets too far ahead of the product, the market eventually catches up. On AI, that reckoning is already under way.