Compliance Chiefs Turn to Cognitive Tech for Surveillance

Combining cognitive computing with compliance know-how could enable surveillance officers to spot warning signs in staff behavior, potentially indicating future problems by reading patterns in data that would otherwise be invisible. It sounds like something out of Minority Report, but chief compliance officers are adamant that this is the future of their craft.

“Coming from a surveillance point of view, I tend to believe that, in the right circumstances, anyone can do something really bad,” said Joe Lodato, global head of compliance technology and surveillance at Guggenheim Securities. “By keying on some of the emotional things that the systems can tell us, that may be an early indicator that we have someone who is about to go off the rails.”

Such information could include, as an example, an employee’s financial health. A bankruptcy or ongoing financial issues could indicate that a member of staff could be prone to making mistakes at work, on the lower end of the scale, or gearing up to commit fraud at the extreme end. Using cognitive technology, surveillance officers can now read these early warning signs and engage in enhanced oversight if they feel it is warranted.

“Compliance tends to want to have all of this data at their fingertips. Now they’re starting to get that data, and get it downloaded so that we can see these patterns and maybe intercede when necessary,” Lodato continued.

The executives quoted in this article spoke on a panel at the RegTech America conference, held in New York on June 14.

Analyzing unstructured data has been a part of sophisticated surveillance platforms for some time, and in the past was often achieved through the use of dedicated complex events processing platforms. Vendor NICE Actimize, for instance, has combined this type of data analysis with its Proactive Surveillance platform for years, while Nasdaq recently partnered with cognitive computing provider Digital Reasoning to enhance its Smarts surveillance platform’s capabilities.

But the breadth of information now available means that surveillance officers have more at their fingertips than just email and chat messages, card swipes into the building and vacation time utilization to flag up problems, and with cognitive technology, the computing power to analyze this information is available. This development has become necessary in part due to the sheer level of data that individuals—and society more broadly—produces on a global scale. Information that might not have been available to watchdogs in the past is now out there.

“We’re doubling the amount of data that we produce and store every two years,” said Chris Vickey, COO for the Americas at Nomura. “Think about what that means and it’s incredible—it means that in the past two years we’ve produced more data than in the rest of recorded history.”

This information can also have real-world consequences. Abel Picardi, chief compliance officer at Bank of China, recalled an incident where the bank ran a background check on a group of interns who had been offered positions at the bank a few years ago. One intern had an offer rescinded over an offensive Facebook post that was published three years before that point.

But it also has benefits. Information that, in isolation, may appear unremarkable to human officers may fit into a pattern that only machines can determine, given the sheer quantity of data that enters and leaves a bank at any time. Scandals related to conduct, in particular, can be effectively tracked by cognitive technology.

“[IBM’s cognitive platform] Watson probably would have picked up Wells Fargo’s problems—based on that data. I’m sure somewhere, someone was complaining about the accounts, and it would have picked up on it,” Guggenheim’s Lodato said.

No Privacy

While such analysis has been the norm for many years in the defense and intelligence industries, it is still developing within financial services. “We started seeing interesting things coming out of the intelligence space within the compliance area four or five years ago—much more advanced surveillance techniques. They’re obviously used to using big troves of data and have been for some time now. I think we’re at an evolutionary stage—we’re certainly not mature yet, and there’s a lot more to do,” said Nomura’s Vickery.

But the use of such technology also brings with it questions over controls, and the line between respecting an individual’s privacy and protecting the firm. For some, such as Jonathan Carroll, managing director for enterprise risk IT, technology and operations at Bank of Montreal, when it comes to bank systems, there is no limit.

“A firm I worked for in the past monitored keystrokes,” he recalled. “That’s the level of surveillance we’re talking about.”

Others are even more blunt in their assessment. “I’m amazed at the things people put on the firm’s e-mail system,” said Bank of China’s Picardi, before summarizing his bank’s attitude toward the issue. “We make it very clear that there’s no privacy.”