A new survey from dbt Labs shows analytics engineering is accelerating faster than its guardrails. The 2026 State of Analytics Engineering report, based on responses from 363 practitioners and leaders, details a widening gap between AI-driven development speed and the systems needed to ensure reliable outputs.

The data indicates a clear skew in priorities. While 72% of teams now focus on integrating AI-assisted coding tools, only 24% are applying similar AI effort to pipeline management, testing, and observability. This imbalance is producing tangible anxiety: 71% of data professionals list incorrect or hallucinated data reaching stakeholders as a top concern.

"Two years ago, most practitioners didn't expect to generate the majority of their code with AI. Today, that's the reality," said Jason Ganz of dbt Labs. He argues the practitioner's role is shifting from writing code manually to constructing the trusted infrastructure that allows agentic workflows to operate reliably at scale.

In response, organizational goals are being reshaped. The priority placed on increasing trust in data saw the largest single-year jump, rising from 66% to 83%. Speed remains critical, climbing from 50% to 71%. In contrast, cost reduction saw only a marginal increase. This suggests teams are seeking budget not just for compute, which 57% report is rising, but for the governance work required to make AI outputs dependable. As the report concludes, treating governance as core infrastructure, not an afterthought, is what separates sustainable impact from uncontrolled acceleration.

Source: dbt Labs Blog