Artificial intelligence (AI) is primed to elevate the finance function, and CFOs and financial leaders are poised to lead the way. But with ample use cases and many places to start, how can you best chart course for your organization’s AI journey? How do you measure and weigh the cost, the benefit, and the risk? And how do you convey the value to your stakeholders?
These were the questions our panel of specialists explored in our DX Coffee Chat webinar, “The CFO and AI: Shaping the future of finance”, hosted by Stephanie Terrill, KPMG in Canada's Canadian Managing Partner, Digital and Transformation and National Leader, Management Consulting, and featuring guests CEO of HorizonX, Steve Suarez, and KPMG’s Global Finance Function Benchmarking Lead for Banking, Aris Kossoras.
Read on for an overview of the key takeaways from this discussion, as well as a summary of recommended actions.
Takeaway #1: AI in finance – Adoption is just beginning
It may seem that AI is being adopted in every corner of every industry, but the truth is many organizations are still in the early stages of adoption. Polling our DX Coffee Chat audience revealed that only 4% of finance professionals have an advanced level of adoption of AI. In comparison, 43% said their adoption was limited, and over a third (38%) had yet to adopt AI.
How far along is your organization in adopting AI for the finance function?
Low AI adoption levels can also be attributed to a lack of formalized AI strategies to unlock the technology’s full and integrated potential and, importantly, tie AI integration to business value. Only 6% of webinar attendees have such a strategy in place, while a majority (58%) revealed they do not. This is in line with what our speakers have generally been seeing in the market.
“Finance professionals are still very much in that initial stage of discovery, assessment, and figuring out the business case for AI," said Stephanie Terrill. “From what we're seeing, a majority do not have a formalized AI strategy or are in the process of developing an AI strategy. They can be assured, though, that this is consistent with where the market is and what we're seeing across the world.”
Does your organization have a formal Artificial Intelligence (AI) strategy in place?
Takeaway #2: Compelling use cases are present, but awareness of them may not be
There is no shortage of use cases for AI in Finance. It has the potential to automate over 90% of transaction processes, streamline the financial close and planning cycles (KPMG, How to elevate finance value through Generative AI, 2023), reduce or eliminate manual tasks, and play a significant role in identifying errors and discrepancies. And, of course, AI/machine learning has the ability to parse massive amounts of internal data to augment and inform an organization’s budgeting, planning, and forecasting activities.
As Aris Kossoras noted, "AI and machine learning techniques can massively augment your planning and forecasting models using both internal data and historical data, in terms of how the business has reacted to changes in the market and external signals that will allow you to project the market forward.”
Speaking to the predictive capabilities in particular, he added. “These types of applications allow the CFO to be able to shape the organization and do ‘what if’ scenario modelling quickly and effectively, making finance ‘directionally right and fast’ rather than ‘precisely wrong and slow.’”
But while there is plenty of value to be gained, based on the polling data from our webinar, finance leaders feel there’s still a ways to go when it comes to their finance function having a strong grasp of AI and Gen AI concepts, as well as the potential applications for their businesses.
Our finance function has a strong grasp of AI and Gen AI concepts and the potential applications in our business?
Takeaway #3: Making the business case for AI entails clarity of communication and realistic expectations
How can CFOs effectively communicate the value proposition of AI to stakeholders? Our DX Coffee Chat speakers shared the perspective that it starts with demonstrating how the technology not only supports productivity and cost reductions in the Finance department, but also directly contributes to an organization’s financial health through strong reporting, planning, and forecasting capabilities.
Following implementation, making an ongoing case for AI requires measuring, quantifying, and communicating the productivity gains and cost savings that can be gained by bringing this technology into the Finance department. This is where establishing and tracking key performance indicators (KPIs) for each use case is advantageous (e.g. efficiency, speed, error rate, etc.).
“Make sure that when you put these measures together for yourself and the rest of the organization that these measures are clear and understandable,” emphasized Kossoras, noting, “And set the proper expectations. These AI’s are trained on humans. We’re human, and we don’t expect to be 100% on everything, so don’t expect your AI to be 100% either.”
Takeaway #4: The key to unlocking AI’s full potential is your people
AI won’t replace the human workforce, but finance teams will require up- and re-skilling to reveal its full potential. This means understanding how to use AI to advance Finance activities, unlock efficiencies, and produce insightful data that drive business strategies.
“The average finance function right now is still predominantly geared around accounting skillsets,” said Kossoras, “Moving forward, though, there will need to be an investment in re-skilling the finance workforce [when it comes to] data disciplines, playing the role of the finance business partner more effectively, and contributing more to enterprise value creation.”
Takeaway #5: Risks are part of the journey
No game-changing technology comes without its risks and challenges. As such, a strong strategy is required to manage and mitigate the internal, external, and ethical risks of bringing AI into finance.
These risks are already on CFOs’ radars. During the DX Coffee Chat, attendees voted data breaches and security issues as their top perceived risk to bringing AI into the finance function, followed by the risks of not having the human skills to effectively wield AI tools, difficulties with integrating AI with existing systems, and regulatory and compliance issues. As with all business transformations, this and other perceived threats can be addressed with effective and proactive risk management.
As a finance leader in your organization, what do you see as the biggest risk associated with adopting AI in the finance function?
The recommended actions from our speakers
1) Avoid rushing in: AI FOMO (fear of missing out) is real, and it can tempt teams to invest in AI solutions that sound good on paper but don’t necessarily add value to the business. Take time to understand which AI applications will bring the most value to your organization.
As Steve Suarez said: “If you’re not doing AI, it can feel like you’re missing out. And when people feel like they’re missing out, they throw money at things or try things that don’t really add value. That’s why the important thing right now with AI is to put a good strategy together and move ahead with a solid approach.”
2) Align on a cross-functional approach for prioritization, tech choices, compliance, and change management: The most successful AI strategies unify both people and processes.
3) Approach the transformation in steps: Define an AI strategy that aligns initiatives, sequencing, and investments with the organization’s strategic priorities. Take incremental steps, measure the results, and go forward at a pace that makes sense for your operations. That being said, don’t wait until your data is perfect (because it might never be).
4) Enable experimentation: Launch pilots, score quick wins, learn, and build momentum. Establish the tools, supports, and guardrails so your team can pilot AI initiatives effectively and with confidence.
5) Consider small language models and their immediate applications: Small language models cost about 1% of the cost to run a large language model and are often the best fit for what finance teams are attempting to achieve.
“Our language model doesn't need to know what [everyone is] doing to solve some of our specific use cases, and we don't have to send it outside of our organization. These models can work on the edge, work with us,” said Kossoras, noting smaller language models can be tailored to specific use cases, generating better outputs, reducing variances and AI “hallucinations,” as well as preventing other big organizations from being trained on your data.
6) Upskill your finance function colleagues to align with the requirements of an AI-enabled organization: Educate your finance team on data disciplines and foster skills that will enable them to leverage AI to contribute to enterprise value creation. Moreover, appoint and empower finance leaders to oversee components of your AI strategy.
Connect with us
Want to bring the full power of AI to your finance function? Our team of advisors is here to help you. Connect with us for an initial consultation on how to streamline and optimize your AI journey.
Stay up to date with what matters to you
Gain access to personalized content based on your interests by signing up today
Connect with us
- Find office locations kpmg.findOfficeLocations
- kpmg.emailUs
- Social media @ KPMG kpmg.socialMedia