Gartner Warns Over 40% of Agentic AI Projects Will Be Abandoned by 2027 Amid High Costs and Hype

More than 40% of current agentic artificial intelligence (AI) projects will be discontinued by the end of 2027, according to a new report from technology research firm Gartner. The firm attributes this anticipated wave of cancellations to the rising costs and the lack of clear business value associated with these projects. Despite widespread interest, many organizations are finding it difficult to justify continued investment in agentic AI systems that have yet to deliver on their promises.

Agentic AI refers to systems capable of acting independently to accomplish defined goals without continuous human input. Major tech companies such as Salesforce and Oracle have heavily invested in this technology, hoping it will improve efficiency and reduce operational costs. However, Gartner suggests that much of the current enthusiasm is driven more by hype than by proven success, with many projects still stuck in early experimental phases.

A growing issue, according to Gartner, is “agent washing”—a trend where companies label basic AI tools like chatbots and digital assistants as agentic AI, even when they lack autonomous decision-making abilities. Out of thousands of vendors, Gartner believes only around 130 genuinely offer true agentic AI solutions. This dilution of the market makes it harder for businesses to distinguish between functional innovation and rebranded tools.

Anushree Verma, a Senior Director Analyst at Gartner, noted that many current projects are poorly aligned with real business needs. She emphasized that existing models generally lack the sophistication to handle complex goals or understand nuanced instructions over time. As a result, companies are realizing limited returns on their AI investments and are likely to pull back.

Despite the challenges, Gartner remains optimistic about agentic AI’s long-term impact. By 2028, it forecasts that 15% of day-to-day business decisions will be made autonomously by these systems, compared to none in 2024. Furthermore, it projects that agentic AI will be integrated into 33% of enterprise software applications by 2028, a substantial jump from under 1% today.

Source: Reuters