Generative AI has made enterprise analytics more accessible to people across organizations, according to MIT Technology Review AI. The next step is agentic AI, which aims to automate entire analytics workflows and take actions based on the insights its systems generate.
That promise comes with a central limitation: the quality of the insights produced by current agents remains a problem, the publication reports. If organizations rely on agents not only to analyze data but also to trigger follow-up actions, the reliability and usefulness of those insights become more consequential.
The report focuses on how the industry can raise the quality of agent-generated analytics as these systems move beyond answering questions and toward carrying out work.