Gartner’s research indicates that many agentic AI initiatives are still in early experimental phases, often fueled more by hype than by strategic planning
More than 40% of agentic AI projects will be cancelled by the end of 2027 as organizations struggle with rising costs, unclear business value and inadequate risk controls, according to a new forecast from Gartner.
Agentic AI refers to artificial intelligence systems that have the agency — within defined guardrails — to go beyond merely augmenting workflows to fully automating them. These systems can take user intent, access relevant data and applications and produce outcomes with minimal human intervention. As MediaTek’s James Chen put it, “Agentic AI takes generative AI one step further… it’s like having an agent living inside a computer — it’s basically a coordinator.”
The concept of “embodied cognition” also plays a role: AI systems can achieve full agency only if they can interact with the physical world. NVIDIA CEO Jensen Huang recently highlighted the company’s push toward “physical AI” during a CES keynote. You can check our more on that here and here.
But Gartner’s research indicates that many agentic AI initiatives are still in early experimental phases, often fueled more by hype than by strategic planning. As a result, these projects frequently stall before reaching production.
“Most agentic AI projects right now are early-stage experiments or proof of concepts that are mostly driven by hype and are often misapplied,” said Anushree Verma, senior director analyst at Gartner. “This can blind organi[z]ations to the real cost and complexity of deploying AI agents at scale, stalling projects from moving into production. They need to cut through the hype to make careful, strategic decisions about where and how they apply this emerging technology.”
A January 2025 Gartner poll of 3,412 webinar attendees found that 19% of organizations reported significant investments in agentic AI, 42% made conservative investments, 8% had not invested at all and 31% were either waiting or unsure.
Complicating matters, Gartner identified a widespread trend of “agent washing,” where vendors rebrand existing AI assistants, chatbots, or robotic process automation (RPA) tools as “agentic AI” without delivering true agentic capabilities. Of the thousands of vendors claiming agentic solutions, Gartner estimates only about 130 actually offer genuine agentic features.
Verma noted that today’s agentic AI models often lack the maturity needed to autonomously execute complex business objectives or follow nuanced instructions—limiting their return on investment. “Many use cases positioned as agentic today don’t require agentic implementations,” she said.
Some companies have already experienced the pitfalls of overhyping agentic AI. In a recent column, RCR’s Sean Kinney highlighted how Swedish payment processor Klarna paused hiring for certain roles and instead deployed AI tools to handle inbound customer service. But Klarna CEO Sebastian Siemiatkowski later told Bloomberg the AI delivered “lower quality” work than human staff. Klarna has since resumed hiring humans for those positions.
“That’s just one example of a company going hard on AI-first labor, then course correcting after the technology didn’t deliver. There are more. And there will be more,” Kinney wrote.
Despite early setbacks, Gartner believes agentic AI represents a major leap in AI capabilities. The firm predicts at least 15% of day-to-day work decisions will be made autonomously through agentic AI by 2028, up from virtually none in 2024. Additionally, it expects 33% of enterprise software applications will embed agentic AI by 2028, compared to less than 1% today.
Gartner advises organizations to adopt agentic AI only where it clearly delivers value or measurable ROI. Integrating AI agents into existing systems can disrupt workflows and require costly changes; rethinking workflows from the ground up may be the better strategy.
“To get real value from agentic AI, organi[z]ations must focus on enterprise productivity, not just individual task augmentation,” Verma said. “They can start by using AI agents when decisions are needed, automation for routine workflows and assistants for simple retrieval. It’s about driving business value through cost, quality, speed and scale.”
