AI chatbots are coming for finance

Hey, have you heard about these things called AI chatbots? We hear the kids are really into them these days—are they cool or are we facing the first iteration of Skynet? (Will the kids get a Skynet reference?)

AI chatbots like Chat GPT by Open AI, its newly announced successor Chat GPT-4, and Google’s Claude, are the latest contenders as the chatbot arms race heats up. Companies like Morgan Stanley, PwC, Duolingo, and Stripe have already started introducing AI chatbots into their businesses and products.

And the chatbots are coming to the finance function, too. Earlier this month, Bloomberg reported that Brex Inc., a financial software company, is testing the Open AI artificial intelligence tool for a service that will answer questions about corporate budgets, spending, and policy issues.

Brex CEO Henrique Dubugras told Bloomberg that Brex’s chatbot will allow finance teams to spend less time answering questions from employees, and focus more on their jobs. Brex is backed by Tiger Global and PayPal co-founders Peter Thiel and Max Levchin, and has been valued at more than $12 billion.

This technology has the potential to be a game changer for businesses and finance. Bain Capital Ventures partner Christina Melas-Kyriazi told Fortune that AI can be expected to revolutionize roles across the financial services sector. “I expect to see more companies leveraging AI and machine learning to help forecast better and more accurately.”

However, while the technology is moving quickly, it may still be some time before it is fully operational and useful for finance departments, according to Glenn Hopper, director at Eventus Advisory Group and author of Deep Finance: Corporate Finance in the Information Age.

But before CFOs succumb to FOMO and rush to invest in AI chatbot tools, they need to carefully consider the hidden costs and benefits, Hopper told CFO Brew.

So here then are some of the potential costs and benefits of AI chatbots for the finance department.

Benefit: It will speed up financial analysis and decision making. The “Holy Grail” of AI chat is being able to “interact with a machine the way that you would a junior analyst,” said Hopper. While today’s ERP dashboards offer a high-level view of a company’s financials, to really dive into the details requires working with financial analysts and that process can take time, according to Hopper. A fully integrated chat bot with full access to your company’s data can answer financial questions and provide analysis without needing to put in a request.

“You can ask questions, just like you would have an analyst and get this information in pretty near real time,” he said. “That’s going to be amazing.”

Cost: Corporate secrets might not stay secret. Right now, your organization’s proprietary information that Chat GPT scans for answers is also being used by Open AI to train the model. Hopper says there is a risk that that information could lead to leaks of proprietary information. That already seems to be the case with Amazon, which found Chat GPT responses that looked suspiciously close to company data.

“I would strongly advise against anyone putting actual company information that they don’t want shared in there,” Hopper said.

JP Morgan Chase and Verizon have already banned the use of Chat GPT, according to the Wall Street Journal.

Benefit: It could streamline the compliance and reporting processes. AI technology could help accounting and finance professionals do manual compliance and reporting tasks like disclosure review, data matching, and drafting accounting memos by up to 10 times faster according to a white paper by Klarity, a document review automation company.

PwC announced that it was partnering with AI startup Harvey, to offer legal professionals AI tools to mine data and streamline “contract analysis, regulatory compliance, claims management, due diligence, and broader legal advisory and legal consulting services.”

It’s not hard to see how that could be useful for finance departments, said Hopper.

“If you have faith in new lease rules, if you have that loaded into a chatbot and you can ask it, ‘How do I need to record these [leases]?” he said. “There’s a template and…it just vastly streamlines the amount of effort it takes to do that.”

Cost: A loss of institutional knowledge and creativity. Just like our reliance on smartphones means that no one remembers phone numbers anymore, automating processes means that there is no one in the organization with a practical knowledge of information.

There is also the risk that AI could cost accountants and finance professionals their jobs, draining organizations of institutional memory.

“The more you automate, the easier it is to say, ‘Oh, well, that’s another area that we don’t need humans to do,” said Hopper. “If you take the most replicable kind of mindless tasks and automate those, there;s a thought that now we can reduce headcount and finance and accounting.”—DA, LR