Traditionally cautious with new technologies, the financial services sector is adapting quickly to the opportunities that AI promises. At the recent Future B2B AI Impact Summit, industry leaders discussed how financial institutions are working to balance the rapid evolution of AI with compliance and competitive concerns.
From blocking to adopting
Generative AI adoption has been faster than any technology adoption in the past 30 years, said AI Impact keynote speaker Melissa McGregor, Managing Director, Deputy General Counsel, and Corporate Secretary at the Securities Industry and Financial Markets Association (SIFMA).
“Four years ago, we were talking about how to block ChatGPT and similar applications. Now firms are talking about how quickly they can roll those out to all of their employees,” McGregor said.
Compliance and back-office operations are seeing some of the fastest adoption rates, she said. AI tools are revolutionizing routine tasks like anti-money laundering monitoring and communication surveillance. They replace random sampling and basic keyword searches with broad analysis across messaging platforms. However, it remains vital to keep a person in the loop, she said. The automation frees finance professionals to investigate genuine red flags.
McGregor recommended using controlled internal AI platforms to limit data sources to trusted material and urged firms to continuously test external vendor tools.
As for evolving sector skill sets, McGregor added, “Questioning and also having some creativity … might be more helpful because [people with a liberal arts background] can question more what they’re reading and perhaps use the applications in ways that were not previously anticipated.”
Infrastructure, shadow AI and risk
Right now, there’s a tug-and-pull as to how much process control is needed, said Joshua Knox, a technical AI researcher and cybersecurity strategist at Horizon Three.
“Everybody keeps treating all of this AI as pilots,” when they should be treating it as mission-critical, Knox said.
As employees integrate tools into daily workflows, operations become dependent on them, but the systems may not have necessary fail safes and could be vulnerable to cyber attacks..
“Everybody has to be thinking about – how are you defending your perimeter?” he said.
Decision paralysis can stall business progress, warned Dani McCormick, Vice President of Product Management for Global Nexus Solutions at LexisNexis. McCormick added that employees frequently bypass internal policies to use personal AI tools.
“No one talks about the fact that they’re using these tools on the quiet. It’s a hidden, shadow AI secret.”
McCormick emphasized that data accuracy remains the single highest operational risk in high-stakes environments like finance. A robust “trust infrastructure” is required that provides full data provenance and strict citations.
“If you’re booking a holiday, the risks are fairly low, but if you’re in the financial services industry and you’re making these tough decisions day to day, the risks are incredibly high,” McCormick said.
Finding the right tool
To reduce risk, it’s vital to pick the right tool for the problem, instead of assuming one-size-fits-all, said Shane Ernest, Senior Solutions Marketing Manager at Camunda.
For example, while agentic AI is valuable for complex, context-heavy tasks, classical business rules and machine learning remain far more predictable and cost-effective for high-speed, high-volume operations like real-time payments.
Ernest warned of financial scrutiny from boardrooms regarding AI investments.
“The CFO and the board want to know what the results are … What did that actually do for the business?”
Focus on clear business outcomes, avoid single-vendor lock-in, and maintain foundational orchestration layers to control and audit what autonomous agents do, Ernest advised.
All three panelists said financial sector AI journeys should start with two or three high-value, repeatable processes rather than widespread overhauls. By pairing targeted implementations with human oversight and continuous governance, financial institutions can safely harness the technology’s benefits while managing its inherent risks.
Listen to the entire AI Impact webinar The Integration of AI in the Financial Services Industry on-demand here.
To learn more about reaching financial services readers or sponsoring other AI Impact webinars, reach out to [email protected]
Subscribe to SmartBrief financial services newsletters here.