AI is washing over organizations and their data environments, putting pressure on data managers to be able to pull data, almost instantaneously, to fuel AI agents and applications. To meet this need, they are increasingly shifting their emphasis and resources to support the development of robust semantic layers—which serve as interfaces that convert and interpret raw data with technical designations into information that can be employed by business users or analytical tools.
This is one of the findings of a two-wave data analytics survey of more than 800 data executives conducted by Futurum Research. According to the survey, at least 44.5% plan to increase spending on semantic layers during the next 24 months, with an additional 14.4% planning to newly adopt. Combined, nearly 59% of enterprises are directing an incremental budget toward a category that is rapidly repositioning from basic business intelligence tooling to mission-critical AI trust infrastructure.
“Our data shows that the ‘data technician’ era is rapidly evolving into the ‘AI shepherd’ era,” said Futurum analyst Brad Shimmin. “Companies are realizing they can’t just dump raw data into a vector database and hope for the best. IT leaders need to shift budget from day-one ingestion tasks to day-two semantic governance.”
Overall, the semantic layer market is projected to undergo a “rocket ship” growth trajectory, doubling its growth rate from 16% in 2026 to 30% by 2031, the research firm has found. This data category is growing faster than other areas, such as data storage and data engineering.
“The forecast indicates that organizations are increasingly recognizing that Large Language Models (LLMs) cannot operate reliably in a business context without a unified ‘dictionary’ to ground them,” the researchers pointed out. “We are witnessing a fundamental re-invention of the Semantic Layer. Historically viewed as a ‘nice-to-have’ helper for business intelligence dashboards, it is graduating to critical infrastructure status. If you want an AI agent to execute a trade or adjust pricing autonomously, it must understand exactly what ‘Gross Margin’ means. Without a semantic layer, you don’t have agents; you have hallucination engines.”
At the same time, business intelligence and reporting growth is expected to slow to 7% by 2030, as underlying logic migrates to the semantic layer. “The value is moving from the visual dashboard to the logical metric store,” the researchers stated. Plus, as agents move into production during the next 2 years, demand for governable, metric- centric data access is on the rise.
Issues cited in generative AI implementations are primarily infrastructure barriers, including integration complexity (29.3%) and the lack of transactional writeback capabilities (24.6%). The survey’s authors forecast that financial resources are shifting away from legacy plumbing, with data integration expected to grow about 10%–12% between 2025 and 2031, compared to more than 20% growth that is anticipated for intelligence infrastructure, which includes the semantic layer and observability.
The survey documents a shift in priorities. For example, “building AI capabilities” and “increasing trust in data” fell 5.6 and 6.5 percentage points, respectively, as top data team objectives. Measurable outcomes seeing greater adoption include new business opportunities (up almost 5 percentage points), SLA attainment (up more than 3.5 percentage points), and project completion (up 3.2 percentage points).
Almost 45% plan to increase spending to build their semantic layers. In addition, 25.4% prioritized investment in the semantic layer for this year, up from 24%.
Trend awareness jumped 7 percentage points to 19%, the survey showed. Several related technology trends gained ground simultaneously, the survey noted, averaging 8 percentage points, led by graph analytics (up 10.8 percentage points), streaming RAG (up 10.4 percentage points), and ethical data governance (up 10.1 percentage points). AI-augmented and agentic analytics remain the top expected trend at 47.8%. “The absence of any decline, however, warrants healthy skepticism; it likely reflects awareness inflation rather than uniform adoption,” the survey’s authors indicated.