685 enterprise CIOs named what blocks AI agent approval: tools, governance expertise, and ownership. Here’s how to move past them, from a Dataiku/Harris Poll survey.
The platform for AI success is trusted by 1 in 4 of the world’s top companies.
New research reveals a growing appetite for business-built AI, alongside a major gap in the tools needed to support it.

Enterprise analytics is evolving into AI-driven decision-making, opening up new opportunities to put your data, models, and expertise to work.

When approved tools slow you down, workarounds can get the job done while making your AI harder to share, maintain, and trust.

Use plain language, visual workflows, or code to create analytics, models, and agents on enterprise data. Test and refine your work in one governed environment, with enterprise controls in place and tools matched to your skills.
Track ownership, preserve lineage, and manage sign-offs as your project develops. Technical and governance teams can inspect your workflows and validate how data, models, and agents connect before approving deployment.
When the business changes, your agents can keep up. Make routine updates yourself within IT-approved guardrails, so small fixes don’t sit in line behind bigger projects.
Get the Global AI Confessions Report: CIO Edition, 2026
Findings from 685 enterprise CIOs on agent visibility, governance, and accountability.