Fact: Agent AI makes people more valuable.
More than 40% of AI projects will be canceled by the end of 2027. This is a forecast from Gartner published in June 2025, based on a survey of more than 3,400 organizations actively investing in technology.
The reason given is not that the agents are not working. That the people who send them make the wrong decisions. “Many agent AI projects right now are early experiments or proof-of-concepts that are too hype-driven and often misapplied,” according to Anushree Verma, senior director analyst at Gartner.
Organizations deploy agents without a clear strategy, without understanding complexity, and without managing to manage what happens when something goes wrong.
In other words, an agent is only as good as the person behind it.
This is very important in marketing. AI agents in sales are real, fast and in many cases necessary. Agents choose the audience. Agents produce content. Agents optimize shipping times, select offers and plan the entire customer journey automatically, proactively and at a level that no human team can match. Skills exist today and are growing rapidly.
But Gartner’s data reveals a warning and marketing leaders who miss it will find themselves on the wrong side of that 40%.
FOMO causes agent failure
The failure rate described by Gartner is not just a mistake. It starts with fear.
Fear of being left behind. The fear of watching competitors move quickly. Fear of being a CMO who didn’t do it when everyone else did. That fear is driving organizations to use agent AI, not because they have a strategy, but because they can’t be last.
The result is agents built on broken workflows. Agents are fed with bad data. Agents operate outside of management structures that keep them aligned with business objectives. Agents issue… the wrong things, in the wrong ways, at the wrong times.
FOMO is not a strategy. And in the age of agency, it’s a costly mistake.
Washing agent
Gartner has identified a widespread trend it calls “agent washing”… marketers are rebranding existing chatbots and automation tools as agent AI without delivering any real autonomous capabilities. Of the thousands of vendors looking for agent solutions, Gartner estimates that only about 130 offer true agent features. Marketing groups that invest in others don’t get agents. They wear automation with an agent price tag. default with the agent price tag.
The results outweigh the wasted budget. Gartner predicts that by 2026, one-third of companies will harm the customer experience by deploying AI prematurely, eroding brand trust and harming both acquisition and retention.
A personal agent who misreads the customer. A content agent who violates compliance. A travel agent who fills a busy customer with offers at the wrong time. These are the predictable results of releasing autonomous systems without human judgment to guide them.
The capture of traders
Gartner’s third prediction is the most telling. The use of GenAI leads to atrophy of critical thinking skills. As a result, 50 percent of global organizations will require AI due diligence.
Half of all organizations are watching their people get bored because AI is always there to take care of them. In silence. Little by little. Until the day the algorithm is wrong and no one can tell.
In marketing, that’s a problem. Marketing requires judgment – the ability to ask not just what the data says, but what it means. Not just whether the campaign worked, but why. Not just to accept the recommendation of AI, but whether it reflects the type, timing and relationship the company is trying to build.
Those questions cannot be sent to the agent. They need someone to check what the machine thinks is right.
The most dangerous marketer in the age of the agent is not the one who rejects AI. He is the one who accepts everything it produces without hesitation.
Agents cannot be trusted to ask the right questions
The agent can improve the offer. It will not ask if it is given the right object.
It can personalize the message based on behavioral signals. It won’t decide that the right move is to say nothing…to give the customer space, to protect the relationship rather than withdraw from it.
It can generate thousands of content variations and test it. It can’t tell the difference between a dynamic message and a connecting message. It’s unheard of when a campaign that does well on data is silently hurting the brand.
It can make the trip flawlessly. It can’t design that reflects what customers really want from this product, at this time in their lives.
These are not limitations that will be resolved with the next model release. They have structure. The AI is trained in the past. One’s job in marketing is to make judgments about what should happen next, even if the data isn’t there yet to support it.
Advertiser as agent manager
The ideal mental model for the agency era is not human versus machine. It is a human machine that brings people together.
That is the basis of Positionless Marketing. For decades, sales teams functioned as an assembly line with handoffs. Positionless Marketing breaks that model by giving marketers three transformative powers: The Power of Data to quickly access customer information for precise targeting and hyper-personalization, without waiting for engineers; Creative Power to create channel-ready assets like copy and visuals, without waiting for creators; and Optimizing Power to run self-optimizing campaigns with automated journeys and testing, without waiting for analysts. Handoffs are completed.
The Positionless Marketer is a hybrid concept that uses AI agents to move beyond traditional positions. Agents manage what used to be waiting for three separate groups, clearing the assembly line. The advertiser is waiting for no one. They think more, moving in a certain direction while keeping human judgment at the center of all decisions made by agents.
This is a promotion, not a replacement. But it comes with real needs. Marketers cannot think systematically, not just operationally. Who can evaluate the output of AI carefully, not just accept it. Who can be held accountable for what agents do on their behalf.
Gartner’s Daryl Plummer put it bluntly: organizations should prioritize behavioral changes and technology changes as a priority. The technology is ready. The question is whether the people in the marketing organization are there.
The window shrinks
The organizations that will win the next decade of marketing are not the ones that send the most agents. They are what build a person’s ability to direct themselves. Gartner’s 40% forecast is not a warning to slow down. It is a warning to do on purpose. The difference between an agent marketing job that adds value over time and one that wastes budget, violates policy, and erodes customer trust is not technology. It is the human judgment that sits upon it.
Marketing teams need to face the realities of the agent AI era: an agent is only as good as the important person behind it.
The opinions expressed in this article are those of the sponsors. Search Engine Land does not confirm or deny any of the conclusions given above.
