More than 40% of agent AI projects will be canceled by the end of 2027 due to rising costs, unclear value for businesses, or inadequate risk controls, according to theGartner, Inc.
Most of the current agentic AI projects are early-stage experiments or proof-of-concept demonstrations primarily driven byhypeand are often misapplied," he saysAnushree VermaSenior Analyst Director at Gartner. This can blind companies to the true cost and complexity of implementationAI agents at scale, preventing the projects from going into production. They need to overcome the hype to takestrategic decisionsand careful about where and how to apply this emerging technology.”
According to a survey conducted by Gartner in January 2025 with 3,412 webinar participants, 19% said their companies made significant investments in agent AI, 42% made conservative investments, 8% did not make any investments, and the remaining 31% are taking a wait-and-see approach or are unsure.
Many suppliers are contributing to thehypewhen they become involved in the "agent laundering"agent washing) – orebrandingof existing products, such as AI assistants, robotic process automation (RPA), andchatbots, without substantial equivalent resources. Gartner estimates that only about 130 of the thousands of AI agents providers are real.
"Most of the proposed agentic AI proposals lack significant value or return on investment (ROI), as current models do not have the maturity and capability to autonomously achieve complex business goals or follow differentiated instructions over time," says Verma. "Many use cases positioned as identical today do not require identical implementations."
Delivered value to the business
Despite these initial challenges, the trend towards agentic AI represents a leap in AI capabilities and market opportunities. A AI agentwill provide new ways to enhance resource efficiency, automate complex tasks, and introduce business innovations, going beyond the capabilities ofbotsautomation withscriptand virtual assistants.
Gartner predicts that at least 15% of daily work decisions will be made autonomously by agent AI by 2028, up from 0% in 2024. Additionally, 33% of enterprise software applications will include agentic AI by 2028, up from less than 1% in 2024.
At this early stage, Gartner recommends that agentic AI be adopted only where there is a clear delivery of value or ROI. Integrating agents into legacy systems can be technically complex, often disrupting workflows and requiring costly modifications. In many cases, rethinking workflows with agentic AI from the beginning is the ideal path for a successful implementation.
To obtain real value of theAuthentic AI, companies should focus on overall company productivity rather than just improving individual tasks," says Verma. They can start by using AI agents when decisions are needed, for automation of routine workflows, and assistants for simple recovery. It is about generating value for the business through cost, quality, speed, and scale.
Gartner clients can read more in the reportEmerging Tech: Avoid Agentic AI Failure: Build Success Using the Right Use Cases.
