Agentic AI is the great current name in artificial intelligence. And it promises to revolutionize the revolution that the AI has meant, although for the moment a good part of companies and users do not see all those benefits that the marketing departments promise. There is also no consensus among analysts that “revolution” promises become reality, although, of course, there will be exceptions.
Gartner, one of the largest world technological consultants, wanted to give his point of view in a report that in summary comes to say that it is much easier to talk about “agents” of AI and promote them, than to build them to work effectively and provide value.
Agentic, just a fashion?
Gartner has predicted that more than 40% of the so -called agency projects will be canceled at the end of 2027. According to the consultant, this technology is too expensive, it has an uncertain commercial value and, often, does not adapt well to real use cases, which often harms organizations instead of helping them. And this is just the beginning.
It is assumed that agricultural AI represents a great advance with respect to current chatbots and services, allowing “intelligent” bots capable of making autonomous decisions in line with the company’s objectives. These agents are promoting an increase in investment in AI and technology companies are taking advantage of this expectation by raising prices for their customers. In theory, AI agents should be able to configure Windows, navigate the web or even write code on the username.
But none of these supposed revolutions is really happening, Gartner warns. Most AI projects with agents remain experiments in initial stages or proof demonstrations, promoted more for advertising than by practical results. These systems are usually misused by organizations that underestimate their cost and complexity, which, ultimately, prevents many projects from reaching production.
Gartner surveyed 3,412 participants of a web seminar in January 2025, asking them how their organizations were addressing the AI revolution with agents. Only 19% claimed to have made significant investments in AI agents, while 42% adopted a more conservative approach. A small minority (8 %) had not made any investment, and 31 % was not sure or adopted a strategy of waiting and seeing.
The stir around agricultural AI is also driven by what Gartner calls “Agent image washing”. Many companies are simply Reboutizing existing technologiessuch as chatbots, robotic automation of processes and virtual assistants, such as the vanguard agriculture. Actually, only 130 of the thousands of the so -called “AI agents” evaluated by Gartner demonstrated genuine agricultural abilities.
«Most of the proposals of AI with agents They lack significant value or of a return of significant investment, since current models do not have the maturity or the capacity of action necessary to achieve autonomously complex commercial objectives or follow instructions nuanced over time »they assure from Gartner.
In spite of inflated expectations and questionable implementations, the consultant leaves margin for success in the future and ensure that agriculture reaches a tangible value if organizations manage to implement it and manage it effectively. According to the Gartner forecast, IA bots will be responsible for at least 15 % of daily decision making by 2028compared to practically zero in 2024. In addition, a third of business software tools are expected to incorporate agency capabilities by 2028, compared to only 1 % in 2024.
