Agents Are Coming for Data (Just Slowly)
InfoWorld, Thursday, August 6th, 2026
LLMs now write SQL reliably, enabling AI agents to take on data tasks incrementally.
AI agents excel at structured, checkable work such as querying data, yet remain underused in that domain. Large language models became reliably proficient at SQL only recently, unlocking opportunities in data engineering maintenance and context management.
Agents can automate mechanical fixes and validate data assumptions through testing, but genuinely proactive insight remains a research prototype.
Teams should prioritize groundwork, context architecture, and infrastructure able to absorb agent query bursts before adoption accelerates.