As artificial intelligence gains greater autonomy, the mandate to understand and audit these systems becomes increasingly urgent. Globally, a lack of explainability or transparency stands as the primary reason users choose to override or disregard AI-driven recommendations. In the Nordic region, where users discard 27.8 percent of AI advice on average, this deficit in transparency is the leading culprit, closely followed by the system’s failure to fully account for contextual nuances.
”As AI moves toward higher autonomy, organizations face a profound new challenge: maintaining trust in systems that humans do not fully understand,” notes Chris Marshall, VP at IDC. ”Our findings indicate that robust oversight, explainability, accountability, and solid data foundations are becoming absolute imperatives for scaling AI successfully.”
Data Quality and Governance
Encouragingly, the Nordic region has improved its scores across all five core pillars used to measure trustworthy AI compared to 2025: data quality and governance, model management and oversight, explainability and fairness, responsible AI policies, and auditability and accountability. Yet, despite these strides, the region still lags behind global benchmarks in every single one of these five critical areas.
The Data and AI Impact Report: The New Economics of Trust is an initiative by SAS Institute, built on research insights from IDC. This extensive global study surveyed 2,699 decision-makers across 28 countries, all possessing deep knowledge of or influence over their organizations’ data and AI initiatives. The Nordic findings are derived from a weighted aggregate of data from Denmark, Finland, Norway, and Sweden, representing 149 respondents in the region for 2026.
Key Takeaways: Five Essential Insights from the Report
- A 15x Dividend on Returns: Enterprises with the most mature frameworks for trustworthy AI are 15 times more likely to report high returns. Specifically, 62 percent of these vanguard organizations achieve at least double the yield on their investments, compared to a mere 4 percent among their less-prepared peers.
- A Resurgent Nordic Landscape: The region’s Trustworthy AI Index staged a massive year-over-year rally, climbing from 31.5 to 58.3—a 26.8-index-point surge that represents an approximate 85 percent increase.
- The Transparency Deficit: A distinct lack of explainability remains the single greatest barrier to adoption, serving as the primary driver behind users overriding or rejecting AI-generated recommendations.
- The Data Infrastructure Bottleneck: Legacy systems continue to hinder progress, with a striking 17.5 percent of global organizations possessing a fully optimized data infrastructure.
- The Agentic AI Trust Deficit: Trust levels vary significantly by architecture. Globally, confidence stands at 76 percent for generative AI, but drops to 66 percent for agentic AI systems capable of autonomous execution.
