Gartner predicts that 40 percent of AI agent projects will be cancelled by 2027 due to high costs and poor returns on investment, with many projects in early stages and driven by hype.
Most AI agent projects are experimental and often misapplied, with organizations underestimating the cost and complexity of deploying AI agents at scale.
Many so-called AI agents are actually repackaged versions of existing tools like chatbots and robotic process automation, with only a small fraction having true agentic capabilities.
While some projects labeled as agentic lack significant value, Gartner forecasts that AI agents could autonomously make 15 percent of day-to-day decisions by 2028 and be included in 33 percent of enterprise software applications.