Semantics is Power in AI: Gartner Predicts 60% of Agentic Analytics Failures

"In the AI world of 2026, the maxim 'Knowledge is power' could be updated to 'Semantics is power,'" says Tony Baer in SiliconANGLE & theCUBE. AI agents don't apply judgment to questionable data the way human analysts do. They don't recognize that "revenue" means gross in one system and net in another, or that last quarter's churn metric was quietly recalculated midway through the year. They reason from the context they're given, so when that context is absent, the outputs are wrong in ways that are hard to detect and impossible to contain. According to Gartner, by 2028, 60% of agentic analytics projects relying solely on MCP will fail due to a lack of a consistent semantic layer. The semantic layer needs to be not only a data team priority but also an organizational one. We built a practical guide for how to audit, build, and govern an agentic semantic layer that scales with your agents, eliminates hallucinations, and gives every user answers they can trust.

A consistent semantic layer is becoming essential infrastructure for reliable enterprise AI. Agents can only reason from the definitions and context they receive, so unclear metrics and conflicting business terms will scale errors just as quickly as automation scales insight. Strong semantic governance is therefore not merely a data issue; it is a foundation for trust, accountability and sound decision-making.

Like
Reply

To view or add a comment, sign in

Explore content categories