Alteryx Survey Quantifies Enterprise AI Ownership Split
A No Jitter brief on an Alteryx survey of 1,400 IT leaders found broad support for IT-business collaboration, but responsibility for strategy, delivery, data access and production remains split.
Enterprise AI programs are running into an ownership problem even when leaders agree that technology and business teams need to work together, No Jitter reported from an Alteryx study of 1,400 IT leaders.
The survey found 71 percent of respondents viewed collaboration between IT and business teams as the most effective strategy for AI success.
Agreement on the model did not settle responsibility for the work: strategy, delivery, requirements, data access and daily production each drew different preferred owners.
IT and technology leaders led the strategy question with 37 percent support, while 28 percent of respondents put that role with executive leadership.
The delivery side leaned further toward technology teams, with 38 percent saying IT or tech leaders should manage and deliver AI and 21 percent choosing data and analytics teams.
Requirements moved closer to the operating units that use the systems.
Business units were selected by 30 percent of respondents to define AI requirements, compared with 21 percent who chose IT or technology leaders for that responsibility.
Day-to-day production created the clearest break from centralized technology control.
For AI and agentic AI in production, 66 percent of respondents placed management at the line-of-business level, suggesting that operational teams expect to own systems once they are embedded in regular work.
Data access remains a separate constraint.
IT or data teams still carried responsibility for access to the information AI runs on in 38 percent of responses, while only 18 percent of business users had full self-service access to cloud data.
That gap limits how independently business units can operate AI workflows.
Governance adds another boundary.
The study found 34 percent of organizations worried about employees using shadow AI tools without IT approval or knowledge, leaving collaboration tied not only to adoption speed but also to oversight.
IDC enterprise AI strategies vice president Alessandro Perilli framed the ownership tension around different incentives and operating speeds between CIO offices and business units.
The survey leaves AI success less as a tooling question than as a division of work among teams that control strategy, data and production.
