Over 40% of agentic AI projects to end by 2027: report

A report from Gartner predicts that over 40% of agentic AI projects will be discontinued by the end of 2027. The trend is driven by high costs and a lack of demonstrated business value.

Agentic AI systems are designed to autonomously achieve goals and take actions.

While companies like Salesforce and Oracle have made significant investments, many products are simply rebranded AI assistants or chatbots. This misleading practice is called “agent washing.”

Of the thousands of vendors, only around 130 offer true agentic AI solutions.

“Most agentic AI projects right now are early-stage experiments or proofs of concept driven by hype and often misapplied,” said Anushree Verma, senior director analyst at Gartner.

Current models often lack the maturity to handle complex business goals or nuanced instructions.

Despite this, Gartner forecasts that by 2028, at least 15% of everyday work decisions will be made autonomously by agentic AI. It also expects 33% of enterprise software applications to incorporate agentic AI by 2028, up from less than 1% in 2024.

🔗 Source: South China Morning Post

🧠 Food for thought

1️⃣ AI’s recurring pattern of hype-to-disappointment cycles extends to agentic AI

Gartner’s forecast aligns with AI’s historical trajectory of enthusiasm followed by reality checks, a pattern documented since AI’s inception in 1956.

IBM notes that AI has repeatedly moved through phases of hype, disappointment, and resurgence, similar to other transformative technologies like electricity 1.

This pattern is validated by broader AI implementation statistics, with research indicating 75-85% of AI projects fail to deliver expected ROI or reach minimum viable product stage 2 3.

The “agent washing” phenomenon Gartner identifies reflects a similar pattern seen in RPA (Robotic Process Automation), where Ernst & Young reported 30-50% of initial projects failed despite significant investment and enthusiasm 4.

Organizations rushing to rebrand existing AI capabilities as “agentic” without addressing fundamental implementation challenges mirrors previous technology adoption cycles where marketing outpaced capability.