REPORT

Harnessing Unstructured Data for AI Innovation: 2026 BARC Study

Most organizations are running AI projects on an incomplete picture. According to 2026 research from BARC, 70% of data and AI leaders report that less than half of their unstructured data is discoverable and usable for AI. Harnessing Unstructured Data for AI Innovation surveys 225 data, analytics, and AI leaders to reveal where AI programs stall, what separates mature adopters from the rest, and the four steps organizations need to take to close the gap.

What you’ll learn

• Why moving AI into production brings data quality, skills gaps, and security into sharper focus 
• What the gap between perceived and actual governance readiness looks like in practice
• Why 70% of organizations can’t access or use the majority of their unstructured data for AI, and why human validation still ranks as a top readiness measure
• How to think about building a platform-independent semantic layer that can access, refine, and consume all data types across platforms 
• The four concrete recommendations BARC gives data leaders for turning unstructured data from a liability into a reliable AI input