How can CFOs leverage generative AI to provide best-in-class “Finance as a Service”? Rahul Malhotra gives his take.
Malhotra is CFO of Charlotte, North Carolina-based Integreon, which provides creative, business and legal outsourced services to corporations and law firms across the globe. He spoke with StrategicCFO360 about what AI makes possible, the risks involved—and how to prepare.
How is generative AI impacting the field of corporate finance? Is it going to be the next frontier—or is it overhyped?
It’s a foregone conclusion that gen AI is the next major tech frontier impacting many, if not most, aspects of business. To understand its impact on finance, we need to understand how the finance function has evolved over the last decade. I refer to this change as finance being a front-view driver as opposed to a rearview one.
Finance is no longer a support function ensuring accounting hygiene and compliance with statutory laws. It has metamorphosed as an independent service offering, “Finance as a Service.” In this context, CFOs are grappling with the potential of gen AI in aiding them become efficient, effective and nimble for both internal and external stakeholders.
Finance as a function—replete with homogenous and repetitive activities—is ripe for technological adaptation. A recent McKinsey report from June 2023 says that 75 percent of the value that gen AI will bring will fall across customer operations, marketing and sales, software engineering and R&D, but that does not preclude finance from realizing benefits as well.
CFOs have been looking into use cases for gen AI, which include analyzing competitors’ earnings report, predicting analyst questions for earnings call. These are baby steps for finance, but it’s a start nonetheless.
What are the risks and challenges associated with adopting gen AI in the finance function?
Gen AI has great promise, but savvy risk management is also a must, given some of the known risks associated with the technology, including:
- Human bias. Gen AI feeds and trains on human input, which can lead to biased models and results that affect the underlying logic on which the AI platform is being built.
- Data security and privacy. This has always been the Achilles’ heel whenever it comes to technological advancement and the creation of newer models. Privacy concerns around technological integration are often debated at length and predate gen AI. Building tools to ensure robust data privacy will be critical in generating confidence among the public and will also expedite acceptance.
- The confidence level of gen AI output remains a challenge and concern. Objective algorithms may be fine in terms of acceptance, but that’s not what gen AI is supposed to do—gen AI by nature is supposed to solve for subjectivity, and that’s where it’s still evolving. It’s prudent for finance leaders to have an additional layer of caution on the output these tools generate.
- Skill availability. The finance function is facing a scarcity of talent around accountants but with gen AI, the shortage of talent is going to become acute, as the skillset for the finance function is changing. Hence, CFOs are struggling between build or buy talent. The finance team of the future will include not only CPAs and accountants but also data scientists.
What are the benefits and opportunities that gen AI can bring to the finance function?
CFOs have been constantly striving to break the mold of being viewed as bookkeepers, to becoming true strategic partners in decision-making. To accomplish this, they need technology to help them gain mental bandwidth, and it can also aid their teams by permitting them to lighten the focus on repetitive tasks and redirect that focus to analytics and insights. As such, gen AI can help primarily by automating and eliminating many of the repetitive tasks, thereby increasing the speed of delivery and data accuracy.
Speed, especially, is critical, as internal stakeholders expect detailed and insightful analytics in real time. Gen AI will facilitate better and accurate forecasting, help close the books faster, in turn facilitating faster and better data analytics.
CFOs are in an enviable position in that they oversee the vast amounts of data that pass through books of accounts. Leveraging AI to be able to manipulate and utilize all the data combinations will help position finance as a valued resourceful business partner. Secondly and consequently, gen AI will help cover the scarcity of accounting personnel that industry is experiencing by the work that gets offloaded to tech solutions.
What are the key steps for effectively integrating gen AI into existing finance processes and systems?
Integrating AI is a transformative journey and like any other initiative—it needs thorough understanding and detailed planning. Key steps to ensure effective, efficient and successful integration include:
- Clarity of understanding. Most technological initiatives fail at the start if there is not a clear understanding of the objective you intend to achieve. You need to understand and identify specific areas where gen AI can help—such as automation of manual tasks, content creation for annual reports and earnings call transcripts, analytics report generation and fraud detection.
- Compatibility and accessibility of data to AI. As a next step, we need to assess the relevance and accessibility of financial data and ensure that it’s compatible and able to handle the inclusiveness of AI technology—and can produce results effectively. Getting the data sets ready, cleansing and making them ready for AI adoption is extremely critical.
- Change management. Accepting change is always tough to accomplish, but the adage “change is constant” has never been more pertinent than in today’s environment. Communication is the key to ensure effective adoption and acceptance and to avoid any resistance in the future.
- Identify AI solutions. Identifying the actual solution or technology. CFOs must conduct extensive research to identify the right solution which matches the requirements and budget.
- Implementing a PoC. Start the implementation by doing a pilot or proof of concept to test the solution and its effectiveness.
The process does not culminate here, as even after a full-blown implementation one needs to constantly monitor the technology and work on continuous improvement and refinement.
