Beyond the difficulty of evaluating a finance organization's return on investment in AI, there are other factors holding back CFOs moving forward.
- Pressure to modernize, forcing us to make wrong choices. Many CFOs are just beginning to understand the difference between predictive analytics and predictive AI, let alone how these tools can help improve the productivity and efficiency of their finance teams. There are very few use cases for AI within financial organizations, so it does not inspire confidence.
- Data Accuracy, Security, and Privacy. CFOs use predictive AI tools that allow finance teams to extract insights from suboptimal data and generative AI tools that can inadvertently expose sensitive performance data to competitors, posing financial stability risks. I'm concerned about what you're doing.
- Bad output, the aforementioned hallucinations. The results of an AI model are only as accurate as the underlying data. If the data is of poor quality, poorly trained, corrupted by things like bias, or if the algorithm is poorly constructed, incorrect assumptions can fuel bad decisions.
- Employee needs and their impact. In the financial industry, skilled technical talent is needed to develop, use, and vet AI tools and solutions, and this talent is in very short supply.
- Slow integration of AI into applications by top-tier financial software vendors. As one CFO said: “If they’re willing to take the time, so should we.”
Early stage
StrategicCFO360 reached out to consultants from McKinsey, Deloitte, and EY to learn more about the obstacles that are slowing the adoption of these tools and solutions and, more importantly, how to overcome them. Four finance executives also commented on their own interests and challenges in AI.
All interviewees agreed that AI is a way for CFOs and finance departments to further reduce the time they spend crunching numbers and strategically advise colleagues in other parts of the company about what the numbers mean for their departments. thinking about. Using machine learning algorithms, both predictive and generative AI can aid in this transformation effort.
For example, predictive AI leverages historical data to identify patterns and trends that can help predict future events. Generative AI, on the other hand, generates new data from existing data and makes other predictions across the value chain of the finance function. “CFO clients who are doing this successfully can now use generative AI to create cash flow forecasts in hours, versus two to three weeks,” says EY Financial Accounting Advisory Services. said Miles Corson, Global and Americas Strategy and Market Leader.
Despite this promise, the use of AI in finance is difficult for many finance executives to navigate. The challenge is the need to understand the interrelationships between different AI solutions, says Ankur Agrawal, a partner in McKinsey's corporate finance and health practice. “Gen AI is not alone in being billed as a changer in the financial industry. The power of technology to transform the finance function is a combination of automation, predictive AI, and generative AI, not one or the other,” he said. explains.
Asked for an example of how different technologies can come together to deliver value to the finance function, Agrawal cited the use of AI in accounts receivable and payable. “Assuming an automated AR and AP process, we can use generative AI as an accelerator, segment the analytics generated by predictive AI, and automate different email templates to different customers at different times. ,” he says.
New customers who are a few days late on their payments will automatically receive a specially tailored email alert, while long-time customers in the same situation with very favorable terms and conditions will not receive a special email alert until several months later. You may not receive a notice of late payment. “If a month goes by without a payment, our generative AI tool will create a new set of emails to send to different customers,” Agrawal says.
“The benefit of combining all three technologies is that it is iterative,” he says, explaining that the data generated by successive machine interactions provides unique insights into customer payment behavior. “Receiving early warning signals about customers with potential credit issues improves cash flow forecasting by better predicting the timing and amount of cash inflows, outflows, and balances.” he says.
numbers game
While there are benefits for CFOs in being a pioneer in new technology, they are often outweighed by the drawbacks of rushing ahead as a pioneer to outdo competitors. “The CFOs I talk to spend a lot of time thinking about AI to understand their use cases,” said Steve Gallucci, global CFO program leader at Deloitte, adding that his conversation , he added, suggests that “most people would be very happy to become fast followers.” ”
Galuchi's colleague Ranjit Lore, Deloitte's US finance and performance leader, agreed. “It's not that CFOs are wary of AI investments; they just don't want to be at the forefront of this space,” he says. “To understand the risks, we need to look at the broader ecosystem of AI use in finance. Currently, they have to look at the broader ecosystem of AI use in finance. They have serious concerns about sexuality. But that doesn't mean they aren't interested.”
This interest has not yet translated into large-scale investment. “Every CFO I talk to is under tremendous pressure to modernize finance through technology, because almost every financial process can benefit from AI technology in some way.” says EY's Colson. “But the reality is a little different than the promise.”
Many CFOs have invested in predictive analytics, but this solution requires human interaction to enter and query data, identify patterns and trends, and test assumptions. Fully autonomous predictive AI eliminates the human presence. That's a problem for many finance managers.
“CFOs are concerned about the data quality of the output, because the numbers need to be accurate all the time, not rarely or here and there,” says Corson. “Over the past year, we have seen generative AI output that cannot be tied to a specific set of numbers or analyses.” To ensure accuracy, financial outcomes produced by AI solutions must needs to be auditable, he says. “Once this happens, CFOs will feel more comfortable using the tools.”
Gallucci agreed. “A CFO's stewardship role requires accurate reporting. No CFO is ready to hand that over to a machine or model that can produce misleading results,” he said. say.
This fear undermines the cost-benefit analysis of the value of AI in finance. “This is a fairly large investment. It's not something you can just do a pilot project and see if it works,” Lore says. “Issues such as trust, control, and accountability for the consequences of hallucinations need to be resolved first.”
Mr. Agrawal expressed a similar view. “The fact that AI gets 'choked' when it makes the wrong decision is a big problem,” he says. “It's a combination of fear of the unknown and not being defined or accountable. It's not just about technology. It's about people and process. CFOs understand that better than others. doing [business leaders]”
It's not just finance executives who are hesitant to adopt AI. Colson noted that many of the best software providers for finance departments have been slow to incorporate generative and predictive AI into their applications. “Many of his CFOs who wanted to introduce AI into the financial sector had to build their own solutions,” he says. “We're starting to see some vendors embed AI in their products, like his BlackLine's business-to-business AI solution for predicting reconciliation issues and trade failures, or HighRadius' Here is an example of his AI-powered invoice processing solution.”
Other challenges include a critical shortage of skilled technical talent who can develop, use, and ultimately vet AI solutions to ensure the output is accurate and auditable. “CFOs recognize that they need to build these different capabilities, but right now CFOs are seen as an impediment,” Gallucci says.
tentative advance
The experts' view is broadly consistent with the cautious approach taken by Robertson, Damiani and CFO Ebeling. Damiani's minimal investment in generative AI in Balentine was “productive and helpful,” he says. You should recheck the output to ensure it is correct. And I'm also concerned about data integrity. We think it will pay off in time. ” The asset management company has his 55 employees and $7.5 billion in assets under management.
Asked about the potential use of predictive AI in finance, Damiani said he was interested but had more concerns than optimism. “As his CFO with a very good controller, predictive AI will create bandwidth for both of us to do more strategic things,” he says. “But there was nothing on my radar that didn't raise any red flags regarding data integrity and security. We are fiduciaries. We cannot simply take being secure for granted. ”
York Solutions CFO Ebeling is refraining from investing more heavily in AI for the time being, but he is still excited about the opportunities available. “We are looking at bringing predictive and generative AI together to track claims, create financial reports, leverage algorithms, and more to make faster and better predictions,” she said. say. “If we can find a way to bring all of this together, that would be great. It would take me and my team out of the rigors of studying numbers and allow us to make more informed strategic decisions.” This IT The consulting firm has 600 employees and primarily serves Fortune 500 companies in the healthcare, financial services, and manufacturing sectors.
Robertson, chief financial officer at Rogers O'Brien, remains bullish on AI despite his desire to avoid the cutting edge. In addition to leveraging predictive analytics to identify and mitigate project completion issues, finance organizations are using Microsoft's generative AI tool Copilot to search email and MS Teams transcripts for content relevant to the finance function. is being developed. The general contractor, who handles commercial construction projects, has more than 500 employees and 2023 revenues of more than $872 million.
“What we are doing now is preparing for the next step,” he says. “We are collecting data differently than before. We now need monthly project status reports that are more than just a bunch of numbers, using generative AI to translate the meaning of the numbers into words. Before, I had to translate.”
Ultimately, Robertson said, predictive AI will be able to predict issues that will cause future delays in project completion, such as only half the number of windows showing up on a construction site due to a slowdown in the supply chain. I look forward to it. “We know that if all the windows aren’t installed on site on time, the project will be delayed 80 percent of the time,” he says. “If AI could suggest what can be done to avoid this possibility in advance of a supply chain slowdown, that would be a game changer.”
Anne Anthony, chief financial officer of Oberon Fuels, which produces low-carbon renewable fuels, has an advantage over the other three finance chiefs. This late-stage startup is in the process of putting together a financial technology stack that includes both predictive and generative AI solutions. “From a back-office perspective, the department I lead is very lean, with just an administrator, a financial modeling guru, and a part-time clerk,” Anthony says.
She plans to implement an ERP system by the end of the year or early next year, just in time for the company's rapid expansion. Partnerships and investments have been secured with Volvo Trucks, Mack Trucks and Ford, and more than 450 fuel and energy customers are already lining up to purchase the company's products.
“My plan is to go big with more technology than we currently need,” Anthony says. “Given the increasingly strategic role of the CFO, I am very focused on acquiring both predictive and generative AI predictive models that can be used across the financial value chain. .”
When other finance leaders pull the lever on such actions, it is inevitable that their colleagues will soon follow suit. Agrawal said: “Next he AI versions will get better and better at transforming finance and will turn his CFOs thinking about AI into solid users.
Source link
