Investment in artificial intelligence (AI) is witnessing a significant surge among enterprises, with Gartner projecting 2026 as a pivotal year for aligning AI initiatives with key business objectives. As the demand for tangible return on investment (ROI) escalates, business executives and technology leaders are turning their attention to agentic AI, which is seen as a powerful driver for achieving measurable financial results.
The technology sector presents a unique opportunity for AI agents, particularly as IT infrastructure expenses are expected to soar two to threefold by 2030, while budgets remain static, according to findings from McKinsey. Over the past year and a half, tech teams—including engineers, developers, and architects—have increasingly employed AI agents, leveraging them not just for task automation but also for managing and coordinating comprehensive workflows that align with business objectives. However, the reliance on these agents raises concerns about the risks of automated decision-making; thus, organizations must ensure that they can trust these systems to execute tasks reliably and securely.
Despite the enthusiasm for agentic AI, a gap remains in readiness, primarily due to the insufficient business context provided to these systems. The complexity of tasks necessitates advanced reasoning capabilities, which are still in the early stages of development. Human oversight remains crucial for the successful deployment of agentic AI, and as technology teams gain more experience with these systems, it is anticipated that confidence in their usage will grow. Insights from a recent survey of 300 global technology professionals reveal that confidence is particularly high for routine tasks, such as report generation and boilerplate code creation, with significant potential in areas requiring multistep workflows and sophisticated decision-making. Notably, data workflows have emerged as a critical domain where teams exhibit the highest trust in AI agents, particularly in contexts such as data quality monitoring and anomaly detection. These insights underscore the transformative potential of agentic AI in enhancing operational efficiency while emphasizing the necessity of human involvement in the decision-making process.
Source: Agent confidence on the technical frontier via MIT Technology Review