Artificial intelligence is at the heart of finance, transforming the way leaders manage teams, processes, and overall strategy.
At the recent MIT Sloan CFO Summit, experienced chief financial officers detailed how the use of AI is helping them and their teams focus on more meaningful, strategic work, whether it’s forecasting, budgeting, or automating repetitive tasks.
in particular,
- MIT Professor Sloan Interview with Jeff HofmeisterShopify’s CFO talks about how the e-commerce platform approaches AI and experimentation.
- and MIT Professor Sloan interviewed jason childThe executive vice president and chief financial officer of semiconductor company Arm Holdings talked about the importance of having clean data before embarking on ambitious AI projects.
Here are four takeaways for finance teams on AI adoption.
Focus on specific use cases and prove ROI before making large investments
So says there is considerable pressure on organizations to invest in AI even when a project does not have a clear framework. Child recommended trying “some small-scale use cases” to prove the return on investment first.
“The best approach is probably to pick a niche: put something out there, test it, figure out where it breaks,” Child said, citing the example of Arm Holdings. Start there. About 60% of Arm’s revenue comes from royalties on its technology, which it pays out on about 8 billion chips per quarter. The company previously used Excel and other tools to predict loyalty, but now uses AI to create predictions.
“We’re actually using the same approach that’s been used for quite some time. It might be a little faster, a little easier, a little less expensive, and maybe a simpler application to do,” Child said.
Hofmeister said there is no one-size-fits-all approach that companies can use when selecting AI technology. what works Engineering does not function according to compliance and vice versa.
“There are so many different applications and use cases out there. In my department, there is no uniformity for any particular application, with a few exceptions,” Hoffmeister says.
2025 MIT CFO Summit: Powering past, present, and future business solutions
Child advised financial professionals not to underestimate the power of simple automation. “Especially in the accounting world, a lot of plumbing ends up [improved by] automation. “We found that with very good automation technology, we could basically solve a lot of these kinds of problems,” he said.
The benefits of generative AI are not so simple for finance. “LLM is probabilistic. Finance is deterministic. There’s an answer. LLM gives you the highest probability of being the right number, but it doesn’t mean it’s going to be exactly the right number,” Child said. “So we need to understand exactly how today’s LLM works and where it can be trusted and where it can’t be trusted. I think of today’s LLM as pretty much a first draft, but it’s never going to be the answer.”
Artificial intelligence for financial services
Face-to-face at MIT Sloan
Register now
Ensure your data is of the highest quality before deploying more advanced AI and machine learning tools
High-quality data is essential for successful AI and machine learning, but it’s not easy. Dimitris Bartsimas, MIT Sloan associate dean for online education and artificial intelligence, estimates that 60% to 80% of the time spent on data analysis projects is spent on data acquisition and cleanup.
Shopify: Dream big, build fast, grow far
Before you do anything else, ask yourself, “How clean is my data? Do I even have the ability to do even the basics?” [robotic process automation] Or compare one spreadsheet to another?” said the child.
“I think it’s very difficult to get a clean data layer,” he said. But “then you can start doing automation layers, you can do machine learning, you can do a lot of different things.”
Hofmeister acknowledged that CFOs often have to make business decisions with incomplete or noisy data or future uncertainty. “Every decision is obviously made at some point with a certain amount of information and a certain amount of context,” he said. “And almost by definition, things change the next day.”
Whether it’s three months or three years from now, finance leaders will need to rethink future decisions as both data and context evolve, Hoffmeister said.
Leverage creativity and fresh thinking when deploying AI
To get the most out of AI, it’s important to experiment among knowledgeable users.
It says AI experiments are consistent with an economics concept called “user innovation.” “Many of the best technological applications were not developed in the lab.” [but] Rather, it is done by individuals who know their particular situation, their problem, and what they think will yield a positive outcome,” he said.
Often, that creativity lies with younger employees, as long as they have the basic skills and respect for data compliance.
Child said younger employees are better at using AI to solve problems. “If you ask people under 25, [about AI use cases]you get an incredibly creative approach that’s really surprising and very novel. ” he said.
But Reppening cautioned that overemphasizing the under-25 demographic could create a workforce that lacks critical thinking and judgment, which are important skills that workers have in a less automated world.
“in [MIT Sloan]”We teach our students how to create basic financial and system dynamics models, but we’re very aware that most of them won’t grow up to be professional modelers. I always justify this by saying, ‘We want you to be educated, solid, model consumers when you leave here. You need to know a little bit about what’s inside to ask the right questions,'” Repenning said.
of MIT Sloan CFO Summitwas launched in 2003 and brings together financial executives from around the world along with leading MIT faculty. This annual event offers a day of interactive learning and thought-provoking talks about the future of finance, accounting, and business. The summit will include presentations, panel discussions, keynote speakers, networking, and a closing reception. The MIT Sloan CFO Summit brings together more than 400 CFOs and senior finance executives. In addition to the CFO, participants include CEOs, presidents, financial executives, audit committees, and investors. The Summit offers 7.5 CPE credits for full-day participation.
Explore the MIT Sloan CFO Summit LinkedIn Group
