The next battleground in programmatic advertising isn’t identity, it’s infrastructure
Opinion
AI needs environments where data and insights can move quickly, securely and intelligently between different participants in the ecosystem. Does adland have the right plumbing?
The digital advertising industry has always been defined by technological change. However, the current conversation around agentic AI follows a familiar pattern: enormous excitement about these systems’ capabilities, but far less attention to whether the foundations exist to deploy them effectively.
In 2026, practically every ad industry conference presents a vision of a near future where autonomous AI agents negotiate media buys, optimise campaigns in real time, personalise creative and make increasingly complex decisions with minimal human intervention.
However, amid the growing focus on what’s possible, the industry risks overlooking a more immediate question: can the infrastructure that underpins programmatic advertising today support the adoption of agentic AI?
The ad industry has seen enough technological change to know that revolutions are usually constrained by infrastructure, not ambition. For example, cloud computing transformed enterprise software only because the underlying infrastructure evolved to support it. Likewise, streaming displaced broadcast when content delivery networks matured. Even programmatic advertising became widely adopted only once exchanges, real-time bidding, and identity frameworks reached sufficient scale and maturity.
Agentic AI will be no different. The organisations that benefit most will not necessarily be those with access to the most advanced models, but those with the infrastructure required to connect data, technology and decision-making across a fragmented advertising ecosystem.
Shifting foundations
The market is already showing signs of movement in terms of infrastructure evolution.
Supply-side platforms are beginning to explore more modular approaches that let services and capabilities deploy more flexibly across different environments. At the same time, demand-side platforms are pursuing closer relationships with publishers through direct integrations, creating faster, richer connections between buyers and sellers.
These incremental changes matter because agentic AI doesn’t simply need access to more data. It needs environments where data and insights can move quickly, securely and intelligently between different participants in the ecosystem. The ad industry’s existing infrastructure was designed primarily to facilitate media auctions; it was not designed to support continuous decision-making between autonomous systems.
In short, the ad industry needs new plumbing.
Programmatic is becoming an interconnected ecosystem
Agentic AI is now blurring the traditional boundaries of the adtech world. Not so long ago, creative platforms created assets, data providers supplied audience intelligence, publishers made inventory available, supply-side platforms managed access and demand-side platforms executed buying. Each party had their own distinct role and area of expertise.
Today, intelligent systems are involved in planning, optimisation, and execution. AI agents cannot make effective decisions if information remains fragmented or moves too slowly between platforms. Success depends on how effectively these technologies interact.
This represents a significant shift in how competitive advantage is created in the adtech space. Proprietary technology within defined boundaries becomes less important than the ability to collaborate across technologies, partners, and AI systems without unnecessary friction.
There won’t be a single blueprint
No blueprint yet exists to indicate what this future architecture should look like. A variety of tech vendors are pursuing different approaches, making investment decisions complicated for brand advertisers.
Businesses understandably want confidence before committing significant resources. However, waiting for universal standards to emerge carries its own risks. Infrastructure often takes years to develop, test and scale. Organisations that delay may fall behind competitors in establishing AI-enabled workflows.
Rather than betting everything on a single emerging framework, businesses should prioritise flexibility. Open architectures, interoperable platforms and strong technical partnerships will provide agility and resilience as uncertainty dominates the marketplace over the next few years.
Competitive advantage will not come from predicting the winning architecture today, but from building systems that can adapt as standards and technologies continue to evolve.
Infrastructure has become a strategic asset
The technologies that attract the most attention are usually the ones users can see. Large language models, generative AI and intelligent agents naturally get excitement because they promise visible transformation. Infrastructure rarely receives the same recognition because it operates behind the scenes. Still, infrastructure has always determined which innovations succeed and which fail.
Agentic AI may be unlikely to replace planners, traders or analysts overnight, but it will place increasing demands on the systems that connect advertisers, publishers, platforms, data providers and measurement partners. Those demands are already influencing how the programmatic supply chain evolves.
Containerisation, direct publisher integrations and closer collaboration across platforms are not isolated technical improvements. They signal that the industry is preparing for a future where autonomous systems need better ways to exchange information.
History teaches us that infrastructure evolves far more slowly than software, so there’s no time to waste.
By the time agentic AI becomes commonplace across media planning and activation, those best placed to benefit will already have laid the groundwork through interoperable platforms, stronger partnerships and more agile technical architectures.
Mateusz Rumiński is the VP of product at PrimeAudience