Advisors and other financial organizations stand to benefit from artificial intelligence and machine learning applications, as well as regtech and other emerging technologies, as cloud computing becomes more pervasive, but a wide range of challenges remain, industry experts said Wednesday at the Securities Industry and Financial Markets Association Fintech Conference in New York.
When it comes to AI and ML, regulations are an issue and the really difficult part remains translating theory into practice via the implementation and management of these increasingly popular technologies, representatives from companies including Broadridge Financial Solutions and TD Ameritrade said during the panel session “Operationalization of AI.”
Noting that he’s the head of surveillance at TD Ameritrade, Eric Hains said his company has projects “underway right now” with AI and ML. “My particular concern [is] not only does the techniques and technology work for surveillance, but our biggest concern as a firm is bringing along the regulators with us through this process or journey,” he told attendees.
Traditionally, surveillance regulators “expect and understand a rule-based approach,” he pointed out. However, the issue now is that, “as we move to machine learning and AI, one of our biggest hurdles is going to be to bring along the regulators and make sure that they understand the technology [and] they’re comfortable with it,” he explained, adding: “It’s really only then that we feel that they’ll be confident that we still maintain the control environment that we’ve had pre-machine learning.”
Conceding that he’s “not a technologist by any standpoint,” Hains conceded that he faced a “steep learning curve” dealing with ML. Therefore, he said, having a “dedicated group that really does have a deep understanding of what we’re now doing, to be able to not just face the regulators, but to explain it to the rest of us in compliance,” will work to “smooth the path for us” as it goes forward.
Firms, meanwhile, have to figure out where they can apply such technology within their organizations so that it’s most useful. The “rate of change” being seen among AI and other emerging technologies, including blockchain, is happening quickly and “most firms do not have the luxury of” going all-in on all of them, Michael Alexander, president of Wealth & Capital Market Solutions at Broadridge, said. The benefit that a large firm like his has is that it can build a solution with one of the technologies and then “spread out the investments,” he said. He also struck a note of caution, saying companies all too often want to use the “coolest” tool rather than the one that’s most appropriate for their needs.
On a similar note, Hains noted that TD Ameritrade kicked off an AI and ML initiative and there were about 75 projects proposed throughout the firm. The company wound up deciding to move forward on just a handful of them, narrowing them down based on which ones offered “the most bang for the buck” — the ones that could be used in multiple places within the firm, he said.
As in other fields, AI and ML can be used by advisors to perform rote, tedious tasks and free them up to deal with parts of their business that they enjoy much more. But at least some advisors have avoided using AI and ML due to the cost of implementing such technology.
Pointing to a third-party report’s findings, Hains said the companies that are using emerging technologies and investing in them are growing at twice the rate of other firms. So, “it’s really critical” for advisors and other financial firms to start using such technology if they aren’t already, he said, adding: “At the pace of rate and change today, the companies that are not keeping up are really falling behind.”
Data, meanwhile, is the “lifeblood of AI,” according to Randy Guy, chief technology officer for capital markets at Jacksonville, Florida-based fintech firm FIS. After all, “without it, you can’t do anything,” he said. But there’s a growing number of data protection laws out there, including the General Data Protection Regulation, that went into effect in Europe last year, that have to be paid attention to, he noted. If these laws become increasingly tighter, “some of the things that we do today, we may not be able to do in the future,” he said.
In another panel session at the conference, “Effective Use of RegTech to Meet Compliance Challenges,” executives from organizations including BNY Mellon and the Financial Industry Regulatory Authority said such technology can enhance compliance and oversight practices at large and regional firms.
“To me, regtech is all about using innovative technologies to make compliance efficient, effective” and for risk management, according to Kavita Jain, director of the Office of Financial Innovation at FINRA. Technology including ML and natural language processing allow firms “to move away from the traditional siloed review of structured data that’s easily available to a more enterprise-wide review of both structured and unstructured data” from internal and external sources, including everything from newspaper articles to social media that can be collectively used to “proactively identify risk factors,” she said.