A study released at the Gartner 2026 CFO & Finance Executive Conference/Xpo shows that organizations with higher success rates in AI projects within finance are more likely to allocate dedicated personnel roles for the scaling and execution of AI technology. This conclusion comes from a presentation by Marco Steecker, Senior Director of Gartner's Finance Research business, at a session on May 29, which focused on strategies for ensuring finance AI success.

"These organizations have realized that we need dedicated people responsible for AI projects," Steecker said. He further noted that the success of finance AI does not depend on budget investment or tool access, but rather on how organizations structure the execution architecture of AI.

According to research data presented at the session, finance departments report that only about 50% of their AI projects succeed. To better understand why some organizations perform better, Gartner divided respondents into two groups: those with AI success rates below 60% and those above that threshold. The analysis found that differences in outcomes are more related to how AI execution is organized rather than funding or tool levels.

Here are four key takeaways from the session on building a successful finance AI team:

1. Establish a dedicated AI leader

Steecker noted that high-performing finance organizations tend to establish a dedicated AI leader to accelerate AI-driven innovation and ensure alignment with the broader enterprise AI vision. This role is responsible for defining how AI is delivered within the finance function, including the types of teams needed and the processes used to develop and deploy tools. It also involves managing the organization's AI project portfolio, allocating resources, and overseeing investments in short-term AI projects and long-term foundational capabilities.

Another key responsibility is driving AI adoption within the finance function, including overseeing AI literacy programs and supporting change management to ensure broader use of AI tools.

In terms of qualifications, organizations typically look for candidates with more than 10 years of business or technical management experience, at least 5 years of experience leading multidisciplinary teams, and a background in operating model innovation and change management.

"It's hard to find someone who meets all the criteria perfectly," Steecker said. "But in my experience, the latter two (operating model innovation and change management) are much more important than the first two. I've seen many people with sufficient ambition who can succeed in AI leadership roles, as long as they have enough understanding of innovation and the ability to drive change." He emphasized that AI leadership roles do not necessarily require deep technical expertise. He suggested providing structured development time for selected leaders—for example, about one day per week over three months—to complete AI training and build foundational understanding.

2. Build a core execution team

After appointing an AI leader, successful finance organizations begin to build a small, execution-oriented AI delivery team that grows as use cases expand. A key member of the team is the product manager, responsible for driving AI projects end-to-end and accountable for delivery and outcomes. This role is the tactical counterpart to the AI leadership role, translating strategy into execution.

AI product managers need to spend significant time refining products and developing roadmaps, achieved by staying current with industry trends and collaborating on new ideas. They also need to coordinate closely with development teams through regular stand-up meetings and continuously review usage data and end-user feedback to identify improvement opportunities and guide iterations.

Another key component of the team is technical expertise, which may come from data scientists (for analytics-driven use cases) or developers/coders (for automation projects). The specific roles depend on the type of AI work being undertaken.

3. Leverage internal IT talent

When acquiring technical AI and data science expertise, successful finance organizations first leverage existing internal technical staff rather than primarily relying on external consultants. This "borrowing" model immediately enhances capabilities, helping finance teams move faster by leveraging existing expertise. It brings three major advantages: a deeper understanding of shared data systems, standardized approaches to building AI tools, and familiarity with existing centralized governance and IT requirements.

Steecker said that, at least in the short term, relying on internal experts can help finance organizations avoid early technical hurdles, thereby increasing success rates. He mentioned that hiring external talent can be a long-term strategy.

4. Design architecture for scaling

As AI applications in finance increase, team structures often become more specialized. Organizations typically distribute responsibilities across product management, technical development, integration, and end-user engagement. This helps manage a growing portfolio of AI tools while maintaining alignment with business priorities.

This approach reflects a broader principle emphasized at the session: AI should not be viewed as a one-time implementation, but as an ongoing capability that requires continuous oversight and refinement. Steecker said: "Treating AI as a 'side-of-desk activity'—tinkering with it only when there's time—is a mindset that too easily leads to failure. It's this mindset that has ruined many AI projects."