AI Agents are Coming for Advertising, Ready or Not: InMobi’s Abhay Singhai
The advertising industry’s next big customer may not be human. It may not even have a pulse. But it could have a credit card, a shopping list and remarkably little patience for a website that can’t answer its questions.
For Abhay Singhai, co-founder and chief executive of mobile advertising platform InMobi, the rise of artificial intelligence agents promises to change how brands attract customers, sell products and measure the effectiveness of advertising.
“The era of agentic advertising is gonna be all about having agents that understand your brand,” Singhai told Beet.TV contributor David Kaplan at the Beet.TV Leadership Summit with Omnicom Media during Advertising Week New York.
These AI systems won’t merely generate advertising copy or recommend products. They’ll increasingly make decisions and complete purchases on behalf of consumers, potentially turning the traditional shopping journey into a conversation between machines.
Most brands aren’t ready for their robot customers
There’s just one problem. Most online merchants aren’t prepared to welcome these new shoppers.
Singhai said InMobi analyzed more than 3.6 million U.S. storefronts and found that only about 30% could support the full process of being discovered by AI agents and completing transactions through them.
That leaves a considerable number of businesses effectively hanging a “No Robots Allowed” sign on their digital front doors, whether they realize it or not.
“Being agentic-ready is all about having an ability that your offering can be read by machine,” Singhai said.
That means making product information accessible, allowing agents to discover prices and enabling purchases with the consumer’s permission.
With the holiday shopping season approaching, Singhai sees an opportunity for brands to improve their digital storefronts before AI assistants become more influential in purchasing decisions.
Giving AI the keys, but perhaps not the whole key ring
Handing decisions to autonomous software raises an obvious question: How much authority should marketers surrender to machines that sometimes make mistakes with extraordinary confidence?
“Fundamentally, AI is probabilistic,” Singhai said. “We need to give it the guardrails.”
He outlined three levels of decision-making authority. Some tasks can be delegated entirely to AI. Others should require human approval before execution. A third category should remain under human control, with AI limited to offering recommendations.
Marketers will need to establish these boundaries and train their systems accordingly.
It’s a significant departure from conventional software, which generally follows predetermined instructions. AI agents can evaluate options and take actions, but they also require supervision.
In other words, the new digital employee may be exceptionally productive, provided somebody remembers who’s the boss.
From dumb impressions to intelligent conversations
The changes could extend to the advertisements themselves.
Singhai argued that consumers are moving beyond searching for information toward seeking intelligence that can help them make decisions and act on those decisions.
“Ads are already becoming conversational,” he said.
Instead of simply displaying a video or delivering a promotional message, advertising could answer questions, compare options, help consumers discover prices and guide them toward a purchase.
Singhai predicted that advertising will eventually “look and sound way more intelligent than just a dumb impression.”
For an industry that has spent decades measuring how many times someone sees an ad, that’s a substantial change. The objective increasingly becomes what the consumer, or the consumer’s AI representative, actually does afterward.
InMobi wants to connect advertising with outcomes
InMobi is pursuing that vision through Mobius, its AI intelligence system designed to optimize advertising toward specific business results.
Singhai said the company has reorganized its internal technology around delivering outcomes that advertisers want, whether that means selling a product or persuading a lapsed customer to return.
The approach reflects a broader convergence of advertising and commerce. Rather than treating an ad impression, a website visit and a purchase as separate events, AI agents could help connect those activities into a more continuous process.
For marketers, that could make advertising performance more directly measurable, although the effectiveness of such systems will depend on how reliably they operate.
The customer is human, but the shopper might not be
Looking ahead, Singhai urged marketers to rethink the very definition of a consumer.
Brands will increasingly need to communicate with both people and the AI agents acting on their behalf. Decisions could move back and forth between humans and machines throughout the buying process.
That creates a more complicated customer journey, even as the technology promises to make individual transactions easier.
Singhai remains optimistic that the shift will improve advertising accountability, efficiency and return on investment.
For marketers, the immediate challenge is ensuring their brands can be discovered, understood and purchased through AI systems.
After all, it’s difficult to make a sale when your prospective customer’s personal assistant can’t figure out what you’re selling. Especially when the assistant is a computer that never sleeps, never gets distracted and probably won’t appreciate your clever Super Bowl commercial.
You’re watching “The Beet.TV Leadership Summit with Omnicom Media at Advertising Week 2026,” presented by Cuebiq, Databricks, Newton Research, Roku and Teads. For more videos from this summit, please visit this page.
The advertising industry’s next big customer may not be human. It may not even have a pulse. But it could have a credit card, a shopping list and remarkably little patience for a website that can’t answer its questions.
For Abhay Singhai, co-founder and chief executive of mobile advertising platform InMobi, the rise of artificial intelligence agents promises to change how brands attract customers, sell products and measure the effectiveness of advertising.
“The era of agentic advertising is gonna be all about having agents that understand your brand,” Singhai told Beet.TV contributor David Kaplan at the Beet.TV Leadership Summit with Omnicom Media during Advertising Week New York.
These AI systems won’t merely generate advertising copy or recommend products. They’ll increasingly make decisions and complete purchases on behalf of consumers, potentially turning the traditional shopping journey into a conversation between machines.
Most brands aren’t ready for their robot customers
There’s just one problem. Most online merchants aren’t prepared to welcome these new shoppers.
Singhai said InMobi analyzed more than 3.6 million U.S. storefronts and found that only about 30% could support the full process of being discovered by AI agents and completing transactions through them.
That leaves a considerable number of businesses effectively hanging a “No Robots Allowed” sign on their digital front doors, whether they realize it or not.
“Being agentic-ready is all about having an ability that your offering can be read by machine,” Singhai said.
That means making product information accessible, allowing agents to discover prices and enabling purchases with the consumer’s permission.
With the holiday shopping season approaching, Singhai sees an opportunity for brands to improve their digital storefronts before AI assistants become more influential in purchasing decisions.
Giving AI the keys, but perhaps not the whole key ring
Handing decisions to autonomous software raises an obvious question: How much authority should marketers surrender to machines that sometimes make mistakes with extraordinary confidence?
“Fundamentally, AI is probabilistic,” Singhai said. “We need to give it the guardrails.”
He outlined three levels of decision-making authority. Some tasks can be delegated entirely to AI. Others should require human approval before execution. A third category should remain under human control, with AI limited to offering recommendations.
Marketers will need to establish these boundaries and train their systems accordingly.
It’s a significant departure from conventional software, which generally follows predetermined instructions. AI agents can evaluate options and take actions, but they also require supervision.
In other words, the new digital employee may be exceptionally productive, provided somebody remembers who’s the boss.
From dumb impressions to intelligent conversations
The changes could extend to the advertisements themselves.
Singhai argued that consumers are moving beyond searching for information toward seeking intelligence that can help them make decisions and act on those decisions.
“Ads are already becoming conversational,” he said.
Instead of simply displaying a video or delivering a promotional message, advertising could answer questions, compare options, help consumers discover prices and guide them toward a purchase.
Singhai predicted that advertising will eventually “look and sound way more intelligent than just a dumb impression.”
For an industry that has spent decades measuring how many times someone sees an ad, that’s a substantial change. The objective increasingly becomes what the consumer, or the consumer’s AI representative, actually does afterward.
InMobi wants to connect advertising with outcomes
InMobi is pursuing that vision through Mobius, its AI intelligence system designed to optimize advertising toward specific business results.
Singhai said the company has reorganized its internal technology around delivering outcomes that advertisers want, whether that means selling a product or persuading a lapsed customer to return.
The approach reflects a broader convergence of advertising and commerce. Rather than treating an ad impression, a website visit and a purchase as separate events, AI agents could help connect those activities into a more continuous process.
For marketers, that could make advertising performance more directly measurable, although the effectiveness of such systems will depend on how reliably they operate.
The customer is human, but the shopper might not be
Looking ahead, Singhai urged marketers to rethink the very definition of a consumer.
Brands will increasingly need to communicate with both people and the AI agents acting on their behalf. Decisions could move back and forth between humans and machines throughout the buying process.
That creates a more complicated customer journey, even as the technology promises to make individual transactions easier.
Singhai remains optimistic that the shift will improve advertising accountability, efficiency and return on investment.
For marketers, the immediate challenge is ensuring their brands can be discovered, understood and purchased through AI systems.
After all, it’s difficult to make a sale when your prospective customer’s personal assistant can’t figure out what you’re selling. Especially when the assistant is a computer that never sleeps, never gets distracted and probably won’t appreciate your clever Super Bowl commercial.
You’re watching “The Beet.TV Leadership Summit with Omnicom Media at Advertising Week 2026,” presented by Cuebiq, Databricks, Newton Research, Roku and Teads. For more videos from this summit, please visit this page.
Series Videos
