Advertising technology companies, the glut of intermediaries that sit in between the buyers and sellers of internet advertising, are trying to paper over their legacy infrastructure with agentic bolt-ons, but they’re fundamentally built for the internet as it was in 2006 rather than what it will be in 2026 and beyond.
There is a new internet economy being built, where every stakeholder, from publishers through to agencies and brands, will be able to transact directly with one another through lightweight, purpose-built AI agents.
Digital advertising’s creaking infrastructure is overdue a decommission
To understand why agentic represents such a deep rewriting of the economy underpinning digital advertising, first we need a crash course in how the ads that fund online content and services are bought and sold.
On one end are publishers, the companies that own and operate websites, apps, TV channels, and anywhere else digital media appears. To fund their content production and operations, they sell advertising inventory alongside their media. In digital advertising parlance, this is called the sell side. On the other end are the brands and agencies that purchase this inventory – known as the demand side.
Advertising technology companies provide various solutions to facilitate the trade between buy side and sell side. If you’re not immersed in this industry, you would never think about the amount of work that goes into making an ad appear on a website or a video ad play before the show you’re streaming, but it’s a dizzyingly complex supply chain.
It would take the rest of this article to provide even a brief outline of all the intermediaries between A and B, so instead I recommend taking a glance at this diagram. You don’t need to be an advertising technology expert to notice that it’s a bit much. And that’s just for display advertising.
All those intermediaries compete for the chance to take a cut of the trade flowing between buy side and sell side. But due to the complexity of the supply chain, it can be hard to gauge where the money is going or whether the companies taking a cut are providing value.
This opaque “tech tax” extracted from every transaction has been a thorn in the side of publishers and advertisers for the better part of two decades. But now, thanks to agentic AI, they can take control of the supply chain and sweep out everyone who isn’t providing clear and measurable value.
An agentic supply chain gives nowhere for middlemen to hide
Digital advertising is the ideal candidate for automation, because most of it already is. In fact, the term for everything that isn’t bought and sold directly is programmatic advertising, in other words, automated advertising. The problem is, the power to automate the supply chain has been historically held by the intermediaries, while buyers and sellers simply have to take it on faith that they’re doing a good job.
Once buyers and sellers are equipped with their own AI agents, these agents can automate transactions between the two directly, with purpose-built agents handling everything from bidding on ads to verifying the delivery and measuring the results.
Intermediaries can still insert their own agents into the mix (say, an audience intelligence company that can improve the precision of targeted ads) but the total transparency of this massively truncated supply chain means if they’re not adding value, they won’t stay around for long.
Beyond Agentic Bolt Ons
Much of the industry’s response to agentic AI has been to bolt AI agents onto existing DSPs, SSPs and workflow interfaces. While these solutions can automate tasks and reduce manual effort, they do little to change the underlying architecture of advertising. The same fragmented supply chain, duplicated processes and disconnected systems remain, with AI simply making legacy workflows more efficient.
The real opportunity is to rethink the platform itself. Agentic AI should sit at the core of how advertising operates, continuously understanding audiences, context, intent, supply and outcomes, rather than acting as another layer on top of existing technology.
As brands, agencies and publishers increasingly deploy AI of their own, the role of people will also evolve. Humans will focus less on operating platforms and more on setting strategy, defining objectives and governing outcomes, while AI models execute, learn and optimise. Success will depend on how well these AI systems can orchestrate with one another, sharing intelligence and making coordinated decisions across the entire ecosystem.
In that world, orchestration becomes more valuable than automation. The platforms that succeed won’t be those with the most AI features attached to legacy products, but those built around AI from the ground up, simplifying the supply chain and enabling people and AI to work together in a far more intelligent, connected and effective way.
The next platform layer
For all Big Tech’s faults, they recognise where the industry is heading. Zuckerberg’s assertion that brands will one day need little more than a credit card to launch fully automated campaigns on Meta highlights a much bigger shift: AI is redefining how advertising is bought, sold and optimised.
The lesson isn’t that every brand should buy directly from every publisher, or that every intermediary disappears. It’s that every participant in the ecosystem must rethink the value they provide. Simply adding AI to existing workflows won’t be enough.
As brands, agencies, publishers and platforms increasingly deploy AI of their own, the challenge becomes orchestrating those systems to work together intelligently, efficiently and towards shared outcomes.
For more than 30 years, the digital advertising supply chain has accumulated layers of technology designed to solve yesterday’s problems. The next generation of platforms won’t win by adding another layer. They’ll win by rebuilding around AI, with intelligence embedded at their core and orchestration replacing operational complexity.
The future of advertising isn’t defined by who owns the interface. It’s defined by who creates the most intelligent ecosystem, where people set the strategy, AI executes the work, and technology quietly connects buyers, sellers and media in the simplest, smartest and most effective way possible.
