Financial industry executives believe that companies have little choice but to use artificial intelligence (AI) to navigate today’s increasingly complex regulatory environment and faster product development cycles.
“In 2025, there is pretty much no compliance without AI, because compliance became exponentially harder,” said Alexander Statnikov, co-founder and CEO of Crosswise Risk Management. “Think about all the change management that happens with regulations. Now, states will be stepping in. How you stay on top of it?”
It’s a similar story with business product development cycles. “They got much faster. Rather than being a two-year cycle, it’s been a few weeks. Products are being shipped quicker,” Statnikov said. Add the complexity of banking as a service and handling of third-party risk management as areas to manage, and “this makes AI a necessity for success of compliance.”
At JPMorgan Chase, AI enhances compliance and risk management by improving operational efficiencies and reducing paperwork for both employees and regulators, said Terah Lyons, the bank’s managing director and global head of AI and data policy.
One key application has been in fraud detection, she said, where AI has significantly reduced false positives — when a financial transaction is declined because the system suspects it’s fraudulent.
“That’s important from a customer experience perspective … because it means that there are fewer times our customers are having to make calls to try to make corrections,” Lyons said at this week’s HumanX conference in Las Vegas. “It supports enhanced security, cybersecurity, fraud security for the firm.”
Beyond internal benefits, AI also plays a role in external regulatory processes. JPMorgan collaborates with regulators who must process vast amounts of paperwork, and AI helps them operate more effectively. This aids both the bank and regulators, leading ultimately to better risk management and compliance in the industry, Lyons said.
“It supports a more effective, stable global system overall, to have these tools actually in the public sector on the other side of the table, enabling and supporting banks in that aspect of risk management,” she said.
Read more: Compliance Moved from Cost Center to Growth Engine in 2024
Compliance Is ‘Very Inefficient’
False positives are not only prevalent in financial transactions; they are also present in compliance work. This occurs when a human worker thinks a case might not be compliant and spends time digging into it only to discover that it had been fine all along. This wastes time.
AI helps save time by identifying these false positives so human reviewers can focus on the truly problematic cases.
“You’re focusing on the true cases that really require compliance to get involved with,” Anthony Soohoo, CEO of Moneygram, said. “From what we’ve seen so far, it’s made our compliance seem a lot more effective, and I think a lot more efficient.”
Soohoo believes this helps raise job satisfaction, too. “I would assume (employees) are a lot happier because they’re not doing a lot of the tedious work. They are spending most of their time in partnership with AI and … reviewing the ones that actually really need their attention.”
To make staff comfortable with using AI at work, Lyons said JPMorgan uses a practical approach to build organizational trust. “We’ve seen a lot of success from running smaller scale, more internal-facing experiments and scaling those up. If you can provide proof points for value delivery and performance, it makes a huge difference in getting everybody on board.”
Soohoo said the compliance process is “very inefficient” and needs to change.
“The big challenge is keeping your eyes around every element of the process and making sure every document … is done right and reviewed,” he said. “The very first thing that AI unlocks is that it allows, on the compliance side, real-time monitoring of every part of your process.”
Looking ahead, Soohoo said AI will only get better, so companies must get on board. “The quicker an organization aligns to figure out how to solve the problems that come with automation and compliance, as well as AI together holistically, the faster the organizations can be effective.”
It’s no use for companies to think they can avoid the AI wave.
“This transition is similar to what happened with the internet and maybe PCs,” Soohoo said. “The world’s a better place with technology (and staying) involved as a partner versus something we want to try to stop.”
Lyons added that learning how to use and deploy AI is crucial. “Effective risk management practices are going to become … a competitive advantage for firms. The organizations that can get this part right are going to accelerate themselves in so many other dimensions.”
“Customer trust is at the heart of all of that,” she concluded.
