Omnicom is using AI agents to cut out middlemen in media buying
Omnicom just made a move that a lot of people in the advertising industry were expecting, but few thought would happen this soon.
During its Q1 2026 earnings call, the holding company confirmed it is already using AI agents to buy media autonomously, negotiating directly with publishers and bypassing a significant chunk of the intermediary chain that has always eaten into client budgets.
This is not a pilot project on paper. These are real buys, already executed for real clients, through a model called agent-to-agent that connects software agents directly to publisher inventory. To understand the weight of this, it helps to remember how programmatic media buying works today. Between a brand and the ad space it wants to buy, there is a massive lineup of technologies, platforms, and companies charging for every step along the way. DSPs, SSPs, ad exchanges, viewability verifiers, brand safety tools… each one of these links takes a piece of the budget before any ad ever reaches the end consumer.
Omnicom CEO John Wren called this a toll during the earnings call — and made it clear that eliminating those tolls has become a strategic priority for the company. In his words, there are not many large groups pursuing more direct relationships with publishers, but this is a goal Omnicom is actively investing in. What makes this move different from previous promises is that the technology is already working and the results are already showing up. 🚀
The problem the media supply chain has always had
The digital advertising supply chain was built over nearly two decades in a pretty organic fashion, with every new problem spawning a new layer of technology to solve it. When the first questions about viewability came up, verification tools were born. When brand safety became a hot topic, specific platforms popped up for that. When click fraud exploded, invalid traffic verifiers arrived on the scene.
The result was a long, expensive chain full of overlaps, where every participant charges for their presence regardless of how much real value they deliver to the advertiser at the end of the day.
Industry studies, including analyses from the ANA — Association of National Advertisers — have already shown that a significant portion of programmatic media budgets never reaches the publisher, let alone the consumer. A considerable chunk of the money gets stuck in the tech intermediaries that make up this chain. That means for every dollar invested in digital advertising, a meaningful portion simply is not landing where it should. For brands, this is a chronic efficiency problem that directly affects the return on investment of every campaign they run.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
What Omnicom is proposing with AI agents is, in practice, a shortcut along this route. Instead of the budget traveling through this entire toll-heavy path, the software agent connects advertiser demand directly to publisher inventory without needing to pass through each of the traditional intermediaries. This does not eliminate all the technology involved in the process, but it dramatically reduces the number of hands the money passes through before turning into an ad that actually gets shown to someone.
How AI agents work in media buying
The AI agents Omnicom is using are software programs capable of making decisions autonomously, within parameters previously defined by media teams and the clients themselves. Unlike simple automation that executes a fixed sequence of commands, an AI agent can evaluate market conditions in real time, compare available inventory options, calculate projected efficiency, and execute the buy independently — without needing human approval at each step of the process.
Omnicom’s head of AI, Paolo Yuvienco, explained that the infrastructure had already been validated before the company’s investor day the previous month, with real money flowing through to publisher inventory. Since then, the company has executed live buys for multiple clients using the framework, which is built around the Ad Context Protocol, developed specifically for agentic media buying.
In practice, the model works with agents connecting directly to partner publisher APIs, accessing available inventory, and running automated negotiations based on campaign objectives defined for each client. This process happens in fractions of a second and at scale, meaning a single agent can manage hundreds or even thousands of simultaneous transactions without any loss in decision-making quality. What previously required entire teams of media traders and analysts can, with this model, be executed with far less direct human intervention at the operational level.
Worth highlighting: this does not mean media professionals disappear from the process. The human role shifts from operational to strategic. Teams define objectives, calibrate agent parameters, monitor results, and make directional adjustment decisions. The AI agents handle high-frequency execution. This division of labor is what makes the model scalable while keeping strategic control in the right hands.
The role of Acxiom and first-party data
A central point that cannot be overlooked in this equation is the role of Acxiom, the data business Omnicom brought in when it closed the IPG acquisition late last year. Without a robust data layer, agentic media buying would just be a cheaper channel — efficient in execution but limited in the intelligence behind each decision.
That is where Acxiom’s Real ID product comes in. It works as a first-party data backbone, giving Omnicom the ability to segment audiences and target campaigns with precision, without relying on DSP proprietary data or third-party cookies that are in decline. In simple terms, this is the data that makes a shorter supply chain actually valuable. A direct buy with a publisher, powered by high-quality proprietary data, is a combination that delivers not just better cost efficiency but also stronger campaign effectiveness.
This combination of agentic execution and robust data creates a competitive advantage that is hard to replicate quickly. Other holding companies and agencies that want to follow the same path will need to solve two problems at once: develop the tech infrastructure for autonomous buying and secure access to a comparable first-party data asset. Not impossible, but it is a road that demands heavy investment and time.
The conflict with ad tech platforms, in context
Omnicom’s timing with this announcement is no coincidence. The holding companies have spent a good chunk of this year in open conflict with the ad tech platforms through which they funnel their investments. The most visible case involved The Trade Desk, which was at the center of a public dispute with Publicis over a transparency audit — and that tension ended up pulling other holding companies into the conversation, including Omnicom itself.
Underneath the specifics of that fight lies a much more fundamental question: who controls the data, the relationships, and ultimately the margin in programmatic advertising? Buying media through AI agents does not resolve that dispute directly, but it suggests a path to making it irrelevant. If holding companies can buy directly from publishers, with their own data and autonomous execution, dependence on intermediary platforms shrinks — and with it, the bargaining power those platforms wield today.
Wren was pretty direct when addressing the pricing question during the call. He argued that as the chain gets shorter and manual work gets automated, Omnicom’s remaining contribution — strategy, creativity, data intelligence — justifies a higher price. Cutting out middlemen, in that logic, is not just about delivering more useful media for the client. It is about repositioning what Omnicom gets paid for.
What changes for publishers and the market
For publishers, the relationship with this new model is a bit more complex than it might seem at first glance. On one hand, a direct connection with advertisers — without multiple layers of tech intermediation — could mean higher revenue per impression, since less money gets lost along the way. On the other hand, that same dynamic requires publishers to have adequate tech infrastructure to connect directly with holding company agents, which not all of them have today.
Major digital players, like news portals, streaming platforms, and large content networks, are in a much more comfortable position to take advantage of this shift than small and mid-sized publishers. If agentic buying accelerates the consolidation of investment into a smaller number of direct publisher relationships, the winners could gain even more — and those left outside that circle could find themselves increasingly far from the money.
Amir Malik, managing director at consultancy Alvarez & Marsal Digital, said the expectation is that the open web’s share of programmatic will continue shrinking over the next three to five years. According to him, the open web is in structural decline due to multiple factors, including Google AI Mode, Gemini, Claude, and OpenAI acting as primary user interfaces. This trend is also driven by the loss of tracking signals and weak distribution, which pushes budgets toward closed ecosystems with deterministic data, like retail media.
For the market as a whole, Omnicom’s move has the potential to accelerate a transformation that was already underway but advancing at a much slower pace. Other holding companies and independent agencies will need to respond to this shift one way or another, whether by developing similar capabilities in-house or partnering with technology providers that can offer something equivalent. Competition in the agency space has always been heavily based on relationships, creativity, and strategy — but now technological efficiency in media execution becomes a top-tier competitive differentiator.
Intermediaries under real pressure
The intermediaries in the media supply chain, for their part, are facing pressure that cannot be ignored. Companies that exist solely as a link between advertisers and publishers, without adding enough value to justify their presence in the chain, will feel this impact pretty directly.
This is not an immediate extinction scenario, because the transition of an entire market does not happen overnight. Omnicom itself acknowledges that the goal is not necessarily to eliminate DSPs and SSPs entirely from the equation, but rather to reduce the role they play and, consequently, the cut they take. Still, it is a clear signal that the current intermediation model has an expiration date. Companies that figure out how to reinvent themselves and demonstrate real value in this new context have room to stay relevant. Those that insist on the old model will find an increasingly narrow path ahead.
What this move really represents
Omnicom’s decision to go public with this information during its earnings presentation is no accident. A holding company of this size speaks to investors, clients, and the market all at once, and every word in that context is chosen carefully. By confirming that AI agents are already operating and generating real results, the company is signaling that the transformation of media buying is no longer a future vision — it is a present reality, already in execution, already producing measurable efficiency.
From an artificial intelligence standpoint, this is also a concrete, high-impact use case for autonomous agents in an industry that moves billions of dollars a year. Most discussions about AI agents still live in the theoretical realm or in controlled demos. Seeing one of the largest communications holding companies in the world confirm real-world use at scale is a significant milestone, both for the advertising market and for the broader narrative of AI adoption in complex corporate environments.
What is clear in this scenario is that the media supply chain is entering a deep reorganization process, and Omnicom is betting that whoever leads this reorganization will come out ahead for a long time. The efficiency that AI agents bring to media buying — fewer intermediaries, lower costs, more speed, more control — is the kind of advantage that, once established at scale, is very hard to reverse. And for brands investing in advertising, the promise that more of their budget will actually land where it needs to land is, at the end of the day, the most powerful argument any agency can make. 🎯
