The short answer: AI ownership belongs in the business, not IT: wherever value creation happens, such as sales, product or operations. Business teams have the context, and context is a real driver of value from AI. If you want a cost center, give it to IT.
Key takeaways
- In the past, every technology was seen as part of IT. AI shows the most value when it is driven by the business.
- Context drives value creation with AI, and business teams tend to have it.
- Give ownership to the team where value is created: sales, product, operations or another function, depending on the industry.
- Published research points the same way: MIT NANDA, as reported by Fortune, highlights empowering line managers over central AI labs.
The usual assumption
I think most would assume AI is owned by the CIO or CTO, but I am not seeing that as the natural location for AI anymore. In the past, every technology was seen as part of IT. IT would let people use the tools the way IT defined them. Yet AI shows the most value when it is driven by the business, not IT.
What I see
I see IT doing some pretty low-quality projects with AI that have very little impact on the business. Yet when someone in marketing or logistics gets hold of the tools, they do a better job of creating real impact than IT ever did.
This is mainly because context is a real driver of value creation when it comes to using AI, and the business teams tend to have the context and IT does not. Therefore, IT-led AI projects tend to be rather poor at producing results.
What the research says
This section adds published research to my own observation.
- MIT NANDA's 2025 report, as covered by Fortune, found that success came from empowering line managers rather than only central AI labs, and from tools that integrate deeply and adapt over time.
- McKinsey's 2026 survey (The state of AI) found nearly three-quarters of high performers fundamentally redesigning workflows because of AI. Redesigning a workflow is work only the people who run it can do well.
Both studies are about what separates results from effort, not about reporting lines, so I read them as support for my observation, not proof of it. See The State of Enterprise AI, Q4 2026 for how each study defines success.
So where should AI ownership really be?
In the business. Depending on the industry, that changes with every business. Maybe sales, product, or operations. Wherever value creation tends to happen, give that team ownership.
Want a cost center? Give it to IT.
Where IT still matters
A fair counterpoint to consider. Business ownership does not remove the need for security, data access, integration and governance, which take technical expertise. The question here is who owns the outcome. The business owns the result, and IT is a partner on how it is delivered safely.
If you are on a data or analytics team that is treated as overhead, the Profit Center Framework covers how to move from cost center to revenue generator.
Frequently asked questions
Who should own AI in a company?
The business team where value is created: sales, product or operations, depending on the industry. AI shows the most value when it is driven by the business, not IT.
Why do IT-led AI projects underperform?
In my experience, context is a real driver of value from AI, and business teams tend to have it while IT does not. IT-led projects tend to have little impact on the business as a result.
Does that mean IT has no role?
No. This article argues about who owns AI and its outcomes. Security, integration and governance are still work that needs technical expertise, so the business and IT need to work together on them. Ownership of the results is the question.