LAS VEGAS — The future of premium video advertising may be decided in milliseconds – faster than any human trader could ever react.
That’s the promise of agentic buying, which enables autonomous systems to optimize campaigns at the line-item level in real time. The technology represents a fundamental shift in how media is transacted, moving from periodic manual adjustments to continuous, machine-driven optimization.
“This is something that obviously a living human being is not capable of doing today or probably ever,” said Ant McDonagh, Chief Trading Officer, North America, Dentsu, in this video interview with Beet.TV at CES 2026.
Knowledge transfer at scale
Dentsu has been integrating AI across multiple dimensions of its operations, starting with internal knowledge management. The agency is building systems that capture institutional knowledge about what works for specific verticals and clients, making those insights accessible across its global organization.
“We’re now starting to store that in a more meaningful way so that our people across the organization, we’re a very scaled organization, can access that information, transfer those insights, and then put that into the planning process and obviously the activation process,” McDonagh said.
The integration extends to audience discovery, campaign optimization, and technical workflows within activation platforms.
Mid-market clients graduate to video
One unexpected beneficiary of AI integration has been smaller advertisers who previously lacked the resources to execute premium video campaigns effectively. Generative AI capabilities around copy creation and creative optimization are enabling mid-market clients to enter the format for the first time.
According to a July 2025 IAB report, 22% of video ad creatives were built or enhanced using generative AI in 2024, with that figure projected to reach 39% by 2026.
“Because of some of the capabilities within generative AI around copy delivery, copy creation, creative optimization, we’re now seeing some of those clients begin to graduate into this format,” McDonagh said. “It brings new investment into the industry. It brings a lot of new creative opportunities for those clients as well to move beyond channels where they feel safe.”
Dentsu has been building out its AI toolkit to support this expansion. In January 2025, the agency partnered with Adobe to launch Adobe GenStudio Dentsu+, a generative AI-powered marketing ecosystem designed to help brands deliver personalized campaigns more efficiently.
The case for modular platforms
McDonagh argued that open, modular platforms are becoming essential for advertisers seeking greater transparency into how AI drives performance. For the past five to eight years, he noted, value has been tightly bound within fully integrated ecosystems spanning supply, technology, and media.
“I think there’s a real hunger for us to be able to go and get a better view of how AI drives the performance and then obviously to be able to use those inputs to go and then obviously run back into the campaign and deliver better results,” he said.
Modular systems also enable agencies to deploy their own intellectual property into the activation chain. “The ability for us to be able to use that to model, to be able to go and create certain value that is unique to us and then combine that with new data signals that maybe weren’t available in that more monolithic supply chain is really exciting,” McDonagh added.
Humans become policy makers
Despite the autonomous capabilities of AI agents, McDonagh emphasized that human judgment remains critical. He described the emerging model as “setting and supervising” rather than “setting and forgetting.”
“We, over time, will become more policy makers. We will set the models for the agents, we will set the policies for the agents and then we will make the judgment calls on whether a particular approach or a particular strategy is culturally relevant, culturally appropriate,” he said.
Brand suitability and regulatory compliance require human oversight, as does the creative imperative to break out of algorithmic patterns.
You’re watching “Tailwinds of Transformation, a Beet.TV Leadership Series at CES 2026, presented by FreeWheel” For more videos from this summit, please visit this page.
