At the Gartner CDAO Summit in New York last week, almost every breakout session opened with the same questions: “How many times have you heard ‘AI’ today?” “How many times have you talked about ‘agentic’ today?”

There was a note of AI fatigue in the air. CDAOs have been dealing with AI since before it was cool (though for many of us, it’s always been cool!) They’ve spent the last 10-15 years building the predictive models and ML systems that turn data into business outcomes. Sure, AI is THE topic now, but for CDAOs, it’s been the topic all along.

But the stakes are high. Gartner predicts that by 2027, 75% of CDAOs not seen as essential to their organization’s AI success will lose their C-level position. That prediction may sound stark, but I walked away from the summit believing the opposite: The challenges holding back GenAI adoption—governance, security, and complex change management—are exactly the ones CDAOs have been solving for years.

The CDAO Opportunity with GenAI

Gartner’s prediction makes sense when you consider how accessible GenAI has become. Today, GenAI initiatives can emerge from anywhere—marketing, sales, operations—and that ubiquity can disrupt who owns an organization’s AI strategy.

But I see this as a massive opportunity for CDAOs.

According to the HBR article Why Your Company Needs a Chief Data, Analytics, and AI Officer, 53% of organizations are appointing a C-level officer for AI (many by expanding the CDAO role), and 93% say AI is leading to a greater focus and investment in data. The mandates they’re calling out for this role include:

  • Owning the AI strategy
  • Preparing for a new class of risks
  • Developing the AI technology stack
  • Ensuring the company’s data is ready for AI
  • Creating an AI-ready culture
  • Developing internal talent and external partner ecosystems
  • Generating significant ROI

Those mandates should look REALLY familiar to CDAOs. Replace AI with D&A in any of them, and any successful CDAO has already tackled those same problems. Owning the AI data and analytics strategy? Been doing that. Creating an AI-ready data and analytics-ready culture? That’s one of the key challenges they’ve already overcome, according to the Gartner report.

Building a centralized GenAI strategy is the path that successful organizations are taking. McKinsey’s State of AI in 2025 Survey found that AI high performers are over three times more likely than others to say their organization intends to use AI to bring about transformative change. As we all know, transformative enterprise-wide change doesn’t come from individual business units. It comes from a centralized team with a view of the entire company (and its data) and a C-level voice: The CDAO.

Executing your GenAI Strategy

The time is now. McKinsey’s State of AI Survey, Accenture’s front-runners guide to scaling AI, and my in-person sessions at the Gartner CDAO Summit all found the same thing: Most companies are still piloting their GenAI initiatives, and most of those initiatives are failing to provide ROI.

At the summit, a few key strategies stood out:

Get Your Data AI-Ready

You can get a lot of value out of GenAI models today with very little data, which is one of the main reasons usage and adoption has exploded. But so can everyone else. What will differentiate GenAI in your organization is augmenting it with your data, your workflows, your context.

That last point is key. You hopefully already have clean governed data for your existing AI projects, but GenAI requires context, not just data. That means providing models with structured and unstructured sources of business processes, standards, and institutional knowledge. Make sure your data and context are integrated and accessible to the suite of tools you’ll be using as part of your GenAI strategy.

Evolve Your Security and Governance Posture

Data security and governance have been critical for years in this community, and GenAI is adding entirely new aspects to the picture. Shadow AI is a real challenge for every enterprise, so walking the tightrope between protecting your business and providing users with all of the tools and access they want to get their work done will become even more important. Otherwise, you’ll experience AI tool sprawl, low adoption of company-wide projects, and overall worse business outcomes.

Be aware of the security concerns of generative AI tools. When you were building ML models, it was your model constructed from your data. However, with third-party GenAI models, you’re dealing with a lot more black box behavior. The challenges end up expanding your software supply chain security footprint. Make yourself aware of common threats, and work with a trusted partner to protect your supply chain.

On the governance side, you’re going to have more people wanting access to more GenAI tools for more use cases. The earlier you can get your governance standards in place, the sooner you can avoid a “wild west” of everyone talking to data in their own way, or spinning out their own AI apps without proper controls. And since you’ll now be dealing with more than just people, make sure you have a solid policy for governed access by agents and autonomous systems.

Know When to Build and When to Buy

Build vs. buy is a difficult challenge with GenAI. Many commercial offerings are extremely expensive and become obsolete by the time they get through procurement, which kills ROI. But building can be just as difficult, because the correct skillset is hard to hire and hard to retain in this market.

A good way to break this down is to buy for the common cases. These are use cases where you’re looking to optimize standard business processes, cost savings, and productivity. Commercial tools are really strong in those areas and provide the necessary stability, support, and integration with the rest of your infrastructure out of the box. Many are already integrated into the tools your company is using.

Match that with strategic GenAI projects that enable you to innovate or grow your business. This is where your business has unique needs, and your proprietary data provides a real competitive advantage. For your building efforts, look to open source models and tools, and build upon standard and widely adopted frameworks and formats. As we see in our 8th annual State of Data Science report, 92% of respondents are using open source AI tools and models, while 76% indicate that open source will become an even higher priority in the next year.

You should own the tech stack decisions for AI, just as you have for D&A tooling. Select a mix of options that help you achieve more than process improvements and savings. Build an innovation engine for your company.

See You All Next Year!

I want to thank Gartner C-Level Communities, as well as everyone who attended the event, for sharing their ideas, thoughts, and challenges. This community has fought hard for decades to transform their companies’ data into real business value, and they’re ready to do it again with GenAI.

I truly look forward to continuing these conversations next year. See you all in 2026!