“Honestly, our old-fashioned programmatic plumbing… Not made for this age of consumption,” said Sweta Jhunjhunwala, Head of Large Partner Solutions and Channel Partnerships, Google India at AdTech 2026 in New Delhi on March 17, 2026.
she discussed the following Moving to formats like microdramas, AI-driven chat experiences, and large-scale live streaming.
Advertising animation using AI
Jhunjhunwala said Google is “pushing the boundaries of generative AI to truly integrate advertising into content,” adding, “Advertising is no longer intrusive, advertising is truly integrated.” AI that renders ads to appear in existing video content. she said it’s a system Identify “valid surfaces” Within video content, you can place creatives such as “parked cars, digital billboards, and even walls.”
In the demo, ads were displayed embedded within the scene rather than as separate slots. She said that these formats are “Clickable overlay” This allows interaction in parallel with embedded placements. The presentation slides marked the immersive video demo as “experimental.”
AI chat ads use context
Jhunjhunwala said that Google is building chat-based ad formats such as:Understand the context of the conversation” and serve ads accordingly. These ads are “depending on the contextwill be delivered at the moment of. “high will” to the user.
He said the system delivers ads “to the right people at the right time.” In her example, a chat interface can prompt users to purchase ingredients directly based on what they are discussing.
Agent systems move beyond dashboards
Jhunjhunwala said the change is “from a dashboard that calls for action; an agent acting on your behalf;” gave an overview of Google’s agent capabilities. Under experiment and. She described a system built into advertising platforms that displays signals such as:visibility” and “arrival” and a present “Possible solutions directly within your workflow, Expand them too.
Conversational interface for ad interaction
Mr. Jhunjhunwala also said, “AI Help Center Guide” and””Conversation Tools within Google Ad Manager.” These tools allow users to: Query performance data directly using natural languagereturns structured output without manually creating reports.
Predict value and granular delivery
Jhunjhunwala said AI can. “Predict the value of an ad request” before being served. For example, she pointed to live sports as a key use case, which could reach “65 million to 70 million concurrent users.” This marks a shift towards evaluating individual ad requests rather than applying uniform delivery across inventory. As a result, she said, this “enables us to do a lot of things.” More targeted advertising Use AI-powered predictions. ” This also includes determining: When will “more premium creative” be distributed? and how allocate demand Even at large events.
Please also read
