The big picture: Everyone in the tech industry is talking about agentic AI. This supposed next-generation evolution of large language model technology is expected to revolutionize everything, but analysts aren't convinced that future will become a reality. It's much easier to talk about and hype up AI "agents" than it is to actually build one.
Gartner has predicted that over 40 percent of so-called agentic AI projects will be canceled by the end of 2027. The technology is too expensive, has unclear business value, and is often poorly suited to real-world use cases, frequently hindering organizations rather than helping them. And that's just the beginning.
Agentic AI is supposed to represent a major leap forward from today's chatbots and AI services, enabling "intelligent" bots that can make autonomous decisions aligned with company goals. These agents are fueling a surge in AI spending, and tech companies are capitalizing on the hype by raising prices for their customers. In theory, AI agents should be able to configure Windows, browse the web, or even write code on a user's behalf.
But none of these so-called revolutions are truly happening, Gartner warns. Most agentic AI projects remain early-stage experiments or proof-of-concept demos driven more by hype than by practical results. These systems are often misused by organizations that underestimate their cost and complexity, which ultimately prevents many projects from ever reaching production.
Also see: AI Agents Explained
Gartner polled 3,412 webinar participants in January 2025, asking how their organizations were approaching the agentic AI revolution. Just 19 percent reported making significant investments in AI agents, while 42 percent were taking a more conservative approach. A small minority (eight percent) had made no investments at all, and 31 percent were either unsure or adopting a wait-and-see strategy.
The ongoing hype surrounding agentic AI is also being inflated by what Gartner calls "agent washing." Many companies are simply rebranding existing technologies – such as chatbots, robotic process automation, and virtual assistants – as cutting-edge agentic AI. In reality, only around 130 out of thousands of so-called "AI agents" assessed by Gartner exhibited genuine agentic capabilities.
"Most agentic AI propositions lack significant value or return on investment, as current models don't have the maturity and agency to autonomously achieve complex business goals or follow nuanced instructions over time," Gartner's Senior Director Analyst Anushree Verma stated.
Despite the inflated expectations and questionable implementations, agentic AI could still offer tangible value if organizations are able to deploy and manage it effectively. According to Gartner's forecast, AI bots will be responsible for at least 15 percent of day-to-day decision-making by 2028, up from virtually zero in 2024. Additionally, a third of enterprise software tools are expected to incorporate agentic AI capabilities by 2028, compared to just one percent in 2024.
While AI hasn't exactly transformed call centers as promised, its next chapter – with agentic AI – might have a brighter future.