AI in financial forecasting: Adopting a measured approach to improve accuracy and decision-making

August 26, 2025

Drew Maginn

The integration of artificial intelligence (AI) into finance departments is well underway, and forecasting automation has become a priority for many organizations. According to KPMG’s Global AI in Finance report, 82 per cent of Canadian respondents confirmed they are using or piloting AI in their financial practices, with financial forecasting and predictive finance tools leading the way. But widespread adoption doesn’t necessarily translate to widespread success, and many organizations still struggle to define the impact of this technology. To truly reap the benefits of AI and machine learning in finance, organizations must take the time to have the right people, processes and practices in place to ensure it leads to solutions rather than new problems.

HCM Podcast

Produced with Google Notebook LM Using AI Narration

What does using AI in financial forecasting mean?

In financial forecasting, AI can be used to assist with research, data analysis, predictive analysis, and planning. By reviewing data to identify patterns and trends, these tools can predict future outcomes such as revenue projections, expense projections and cash flow.

When used effectively, integration of this type of technology can lead to improved efficiency and accuracy when managing this information. Unfortunately, many organizations invest heavily in AI solutions before taking into account key considerations that can make or break the potential of these investments.

Key considerations when using predictive finance tools

When it comes to financial forecasting using AI, effective adoption is linked to several considerations.

AI opportunity: Opportunities available to an organization to adopt and implement AI into their business practices.

Data and analytics (D&A): Within finance departments, data includes all information that can be accessed and used to assess your financial health, forecast future performance and manage opportunities and risks. Analytics allows organizations to uncover insights to direct their business practices, both now and into the future.
AI maturity: How your organizational need for AI aligns with your current organizational readiness and willingness to implement AI use cases.

AI use cases: Specific ways an organization implements AI to solve problems or improve processes. For financial forecasting, AI use cases are intended to improve efficiency, data management, planning and decision-making.

” Processes should be revised and updated to capitalize on AI, but there is a limit, and effective processes should not be arbitrarily changed to better match with an AI solution. At the end of day, technology needs to fit the process, not the other way around. ”

Walk before your run: AI forecasting best practices for finance teams

Before relying on AI tools to assist with financial forecasting, there needs to be an organizational infrastructure in place to set yourself up for success. This includes adopting best practices like the below.

Driving use of AI tools through process: AI tools are intended to enhance financial processes to benefit your organization. Processes should be revised and updated to capitalize on AI, but there is a limit, and effective processes should not be arbitrarily changed to better match with an AI solution. At the end of day, technology needs to fit the process, not the other way around.

Establishing data governance and ownership: Chief financial officers and senior leaders need to take ownership of organizational D&A, including roles and responsibilities, through a data council with stewards representing key departments. For financial forecasting, this could include representatives from sales, marketing and communications, research and development, human resources and operations. AI can only be used for financial forecasting if the data quality meets a consistent standard. This includes providing finance professionals with access to reliable information from different departments in a format that is compatible with your software and operating systems.

Is each department responsible for providing “perfect” data? According to Gartner Research, organizations adopting AI may need to accept that data volume and volatility will sometimes lead to imperfect data. Accepting this reality, also known as accepting a “sufficient version of the truth,” can accelerate AI use cases by allowing for faster utilization of data for forecasting and decision-making without expecting total accuracy all the time. Strong processes can ensure controls are in place to catch errors or anomalies along the way.
Assessing data access and security: AI tools require data that used to sit in different areas of the organization to be consolidated into one place, sometimes for the first time. Finance professionals need strong technical support and training on security protocols to ensure that all data is kept confidential and protected, especially as it relates to the current and future financial position of the organization. A compliance team should always be kept in the loop when setting up these systems to ensure that breaches do not occur.
Understanding your AI maturity: Many organizations jump into AI use cases for forecasting, whether it be revenue or expense projections, without stopping to think about whether their employees can adapt to the type of change needed to successfully implement the required processes and controls. This includes having employees with a variety of backgrounds (e.g., financial analysis, information technology, accounting, data management) working together to investigate solutions where AI can help, as opposed to resisting change in an effort to maintain the status quo.
Selecting use cases wisely: While there are countless use cases for integrating AI into your forecasting practices, having an approach to help you choose the right ones is crucial. According to Gartner Research, 62 per cent of finance functions do not have guidelines and processes in place that allow them to consistently assess and select AI use cases. Without these in place, researchers recommend organizations start with two key considerations to drive decisions: feasibility and alignment to functional priorities. For example, a volatile market may make inventory projections a priority, while concerns about the financial well-being of your organization might make managing cash flow and maintaining a healthy balance sheet your greater AI integration need.

With AI topping the list of priorities for most organizations, there is no question that all departments, including finance, need to think about use cases that best fit their needs. But before AI tools can become integrated into your financial forecasting practices, there needs to be an organizational commitment to doing things the right way. This includes maintaining strong processes, establishing good governance, managing quality data and retaining a workforce with the skills and openness needed to take on AI use cases.

What are your thoughts on