Editor's note:Georg Glantschnig is Corporate Vice President of AI ERP at Microsoft, based in Redmond, Washington. Cory Hrncirik is Modern Finance Lead in Microsoft's CFO Office. The views expressed in this article are those of the authors alone.

The arrival of generative AI capabilities in finance operations has sparked both excitement and unease among many CFOs. Many finance leaders are leveraging generative AI to streamline repetitive tasks for team members and automate workflows, with the goal of making decisions that can influence the organization's strategic direction.

However, many CFOs report that generative AI experiments have not yet translated into the expected productivity gains.

Deloitte's recentCFO Signalssurvey shows that 70% of CFOs expect generative AI to deliver 1% to 10% productivity gains for their finance functions. CFOs are prioritizing generative AI and expect employees to fully leverage the value this technology creates for the organization.

As the industry moves from AI experimentation to AI implementation, CFOs need to ensure their organizations understand how to best use AI—whether by upskilling existing employees or bringing in talent with generative AI capabilities.

In fact, the CFO Signals survey shows that 60% of CFOs say that bringing in talent with generative AI skills over the next two years is "extremely important" or "very important." However, 61% of CFOs believe generative AI will have little or no impact on their current finance talent model.

Deloitte's data also shows that, as demand for talent with generative AI skills is expected to continue growing, 50% of CFOs say they plan to integrate generative AI by developing existing employees. However, it is not surprising that there is a learning curve in adopting finance-specific generative AI tools.

So, how can CFOs ensure their finance professionals use generative AI capabilities most effectively?

At Microsoft, our finance organization has been on such a journey for years, positioning us well to adopt new capabilities and quickly realize productivity gains. One such tool isMicrosoft Copilot for Finance, which helps finance professionals reduce time spent on repetitive tasks and accelerate access to insights, thereby driving business growth.

With the accounts receivable reconciliation feature in Copilot, our treasury team saves an average of 20 minutes per account, equivalent to roughly a 22% reduction in average processing time costs. In data reconciliation, our financial analysts now spend only about 10 minutes per week on data verification, whereas this process previously took 1 to 2 hours.

Recommendations for applying generative AI in finance teams

The following are best practices distilled from Microsoft's own experience for optimizing generative AI capabilities within the finance function:

  • Address hesitation head-on:Leaders who are hesitant about adopting AI need to recognize that generative AI serves as a "copilot" or assistant for finance professionals, not a tool to replace human work. Outputs must undergo human-level verification. Therefore, finance leaders should focus human effort on tasks that only humans can perform and design user experiences that encourage and help users verify outputs.
  • Educate employees:Finance leaders should clearly communicate to employees the necessity of human verification, the principle of "trust but verify" for outputs, and the core idea that humans remain in charge. This will encourage finance professionals to view generative AI as an empowering tool rather than a replacement threat. When employees feel empowered by generative AI, they will gradually realize its significant impact.
  • Encourage a growth mindset:Successful adoption of generative AI within a finance organization must begin with buy-in from senior leadership. This first requires fostering an environment that encourages a growth mindset. Finance employees need to embrace the idea that, by welcoming new ideas and technologies, they can continuously expand the boundaries of their roles and skills.
  • Embrace the "innovation flywheel":To ensure employees feel empowered by AI, leaders need to create a safe space for functionally diverse and culturally varied teams to experiment safely. Leaders should focus on building an "innovation flywheel"—providing employees with access to new technologies, encouraging experimentation, sharing learnings, celebrating successes, and then continuously repeating this cycle. Both leaders and employees should be confident that the generative AI capabilities supporting their finance tasks perform consistently with limited variance. Encouraging employees to participate in the experimentation process will strengthen their trust in outputs and make adoption more collaborative.
  • Understand that generative AI is not a "silver bullet" for data issues:Leaders should not view generative AI as a magic potion that automatically solves all data problems. Finance leaders need to recognize the importance of getting data in order to maximize generative AI capabilities. Leaders should continuously invest resources in organizing data. Creating rich, well-structured foundational data is critical to optimizing generative AI capabilities in finance.

There is no doubt that generative AI will have a profound impact on redefining how finance professionals work. By putting these principles into practice, finance leaders and CFOs can drive transformation while reaping the outcomes and benefits that support the company's strategic initiatives.