AI Comes for Dealmakers, Wealth Advisors

This article was originally featured in Bloomberg here.

Humanoids in Banking

Jane Fraser says Citigroup Inc. is integrating AI into “wherever it makes sense.” JPMorgan Chief Information Officer Lori Beer says the technology is “actually changing the way we work.” Kristin Milchanowski, chief AI and quantum officer at Bank of Montreal, starts her new book, Return on Intelligence with the line “artificial intelligence has arrived at the intersection of promise and peril,” challenging leaders to absorb rapid change while maintaining human trust. From governance and compliance to cost savings and efficiency, there’s a lot to balance when it comes to AI adoption in financial services. So it’s safe to say it’s one of many critical topics we’ll be exploring at our next Women, Money & Power conference on Dec. 9 in London.

To help set the stage for that discussion, Laura reconnected with Kim Bozzella, an expert who sits at the intersection of technology and finance. Bozzella has a unique perspective with more than 30 years experience in financial services, information technology and consulting. She currently advises leaders across finance and other industries as the global CIO solutions leader at the international business consultancy Protiviti.

AI adoption across financial services has been fast and furious and the industry is a leader in the space, specifically in applying AI to compliance, risk, fraud and other back-office tasks. But when it comes to the revenue-generating areas of wealth advisory and investment banking, there’s a growing recognition among bank leaders that human relationships remain paramount, Bozzella says.

The revenue-generation side is “where they want to put the investment,” she says. “But it’s more complex.”

In investment banking, there’s work currently handled by junior bankers that AI can do such as drafting pitch books. “A lot of those reports will get written and then go to a human that really looks at it and has the view and adds the insights and perspective,” Bozzella says. In the case of wealth management, tasks like preparation, research and analytics might be automated but the adviser-client relationship remains. “Customer intimacy is too important,” she says.

So much for all the hype around robo-advisers.

To get financial firms prepared for this second wave of AI adoption, Bozzella says conversations have shifted away from experimentation and usage to focus on governance, cost control and board oversight, as well as the risks and challenges posed by agentic AI.

With a particular expertise on finance (she spent 14 years at UBS and held roles including global chief information officer of corporate technologies and global head of IT risk management), Bozzella is getting boards to understand that AI adoption isn’t purely a technology issue but a topic that requires broad alignment across departments and a recognition that “it’s not just a question for your CIO.”

Many of her conversations are now focused on how to integrate AI agents — artificial intelligence-powered tools that can handle a variety of tasks, all with minimal human supervision. “It keeps the CISOs, and the risk officers and the CIOs up at night making sure you’ve got those controls in place,” Bozzella says.

Some of the questions she says firms need to think about: How many agents do you have? Where do they sit? What kind of access do they have? What are the controls you have in place for these tools and how do they operate? Can you just bring them into your standard governance framework, or do you need to look at it differently?

“You need to make sure you’re governing and providing the oversight of that, those agents, and keeping an inventory of them,” Bozzella says. The challenge, she says, is “a security and risk management topic. It’s a heavy one. There’s a lot of conversation.”
Going forward, those banks that get the governance right and work out how to build a blended workforce of humans and agents in some of these more revenue-generating areas will be best positioned in three to five years, Bozzella says.

“I do think this agentic piece is going to be really important,” she says. “So, for the banks that have really figured out how to confidently, securely, responsibly manage a workforce that has humans and agents in a lot of processes — not just the heavy operational ones — but really figured out how to make that work in some of the more sophisticated financial services, that’s going be the differentiator.”