Companies are rushing toward agentic AI, but the data underneath isn't ready.

This first-of-its-kind report from a survey of 225 data leaders shows why.

Barc Study Front Cover 4 (900)

The most expensive way to find a data problem is to build AI on top of it.

As companies invest millions in the latest models, they're finding that the unmapped files underneath aren't governed and aren't ready.

The new Harnessing Unstructured Data for AI Innovation report, an independent study by BARC© research, surveyed 225 data leaders across the US and Europe to find out where unmapped files are blocking AI initiatives and what leading organizations are doing about it. 

Get the free report, sent straight to your inbox.
graph@2x

"Only 29% of data leaders say they fully know where their relevant files are for AI." 

BARC Harnessing Unstructured Data for AI Innovation, 2026

Inside you'll find out:

How data quality jumped from obstacle #8 to #3 in two years, and what the shift says about the cost of moving fast on bad foundations.

Chart 1 w border
Chart 2 w border

79% of data leaders are confident they can govern their unstructured data, yet only 36% know where their sensitive data lives.

Why most organizations are still relying on manual processes and individual expertise to govern the data feeding their AI pipelines, and why that's increasingly becoming a liability.

Chart 3 w border

Research sponsored by Ohalo, a trusted partner of Fortune 100 and government organizations

Department of Defense_White (300)
Veolia_White (300)-1
Wood_White (300)
Costain_White (300)
Macnica_White (300)-1
Atlan_white (300)

The research is ready.

 

Find out what’s blocking AI results and what data leaders are doing about it.