CFO Slice

Companies could reduce agentic artificial intelligence (AI) costs by up to 60% and improve AI accuracy by as much as 80% by 2027 if they improve the contextual quality of their data, according to new research presented by Gartner. The firm said many businesses are investing heavily in AI tools without addressing a key problem: underlying data that lacks semantic context, making AI systems more prone to hallucinations, bias, and unreliable outputs. Gartner argued that traditional data models are no longer sufficient and recommended companies build semantic “context layers” into their data infrastructure so AI agents can better understand relationships, rules, and business meaning within enterprise data. Rita Sallam, distinguished vice president analyst at Gartner, warned that failing to address data context could create financial, legal, and reputational risks, while increasing inefficiencies and unnecessary AI spending. The report positions semantic data management as both a cost-control and governance issue for CFOs, particularly as regulators and audit committees pay closer attention to AI-generated outputs used in financial reporting and disclosures.