More regulators talk tough about AI in investing, without slowing its growth

Massachusetts is the first state to join the SEC in probing the potential of AI applications to cause conflicts of interest, but AI poses a much greater danger as well, Gensler continues to warn

Fund boards have one more set of regulators – though not much more detail – to worry about while evaluating the artificial intelligence apps being developed by their funds or advisers.

Securities regulators in Massachusetts announced Aug. 3 that they had sent letters to investment firms, including Morgan Stanley, JP Morgan Chase and E*Trade, asking how each uses artificial intelligence (AI) applications in their businesses. The letters asked different questions of each firm, according to state officials, but all asked how each company makes its AI-enabled systems comply with rules forbidding investment firms from putting their own financial interests ahead of those of investors.

“If deployed without the guardrails necessary to ensure proper disclosure and consideration of conflicts, I am concerned that this technology could result in harm to investors,” according to the announcement issued by the office of William Galvin, Secretary of the Commonwealth, who oversees the Securities Division that regulates corporations operating in the state.

Supervisory procedures are top priority, but the division also asked companies that have deployed AI-enabled applications to provide information about the AI-related information they disclose to investors and asked some firms about marketing material aimed at investors that was created using AI systems.

“State securities regulators have an important role to play when it comes to AI and its impact on main street investors,” Galvin said in the announcement.

Companies that received the letters have until Aug. 16 to reply.

The Securities Division has not revealed whether it would expand the investigation to include other firms or whether it would make disclosures about the use of AI applications mandatory for companies operating in the state.

Massachusetts has not passed any new regulations restricting the use of AI by investment firms, which means the questions sent out by Galvin’s office provide the only new information available to Massachusetts companies about any changes in enforcement policies regarding the use of AI in the investment industry.

Those priorities dovetail with warnings from SEC Chair Gary Gensler, who codified his frequent public warnings about the potential of AI to create conflicts of interest into a set of regulations proposed July 26 that would force investment advisers and broker-dealers to test AI applications for potential conflicts and disclose the steps they take to counteract or neutralize any they find.

“You’re not supposed to put the adviser ahead of the investor, you’re not supposed to put the broker ahead of the investor,” Gensler told the New York Times’ Dealbook Aug. 7.

That hasn’t stopped the market’s rampant enthusiasm for AI, which attracted more than $1bln in venture-capital investments for financial-technology startups during the first half of 2023, during which overall VC investments dropped 49%, according to a July 13 report from S&P Global analysis firm 451 Research.

The biggest areas of investment were focused on digital lending, insurance technology, investment- and capital-market applications, but may be best characterized by Alphasense. The company raised $100m to fund its development of an application that promises to use generative AI – the season’s favorite flavor of machine learning/artificial intelligence – to allow customers to summarize important information from mounds of often-impenetrable financial data easily.

This is the breakout year for generative AI, according to research from McKinsey & Co., which estimates the potential of the new technology to automate as much as 60% to 70% of an average worker’s daily tasks will “change the anatomy of work,” and boost productivity enough to add value of between $200bln and $340bln annually to the banking industry alone, though other financial-services market segments will also see enormous benefits.

Gensler has also praised AI as having the potential to create tremendous growth and improve cost efficiencies in financial services, though that benefit also comes with risk beyond the potential for conflicts of interest, according to a research paper Gensler published six months before he was confirmed as chair of the SEC.

If the growth of AI within the investment business follows patterns common to other disruptive technologies, the industry would likely standardize on a small number of foundational application platforms or analytical models, Gensler argued in his 2020 paper.

That commonality would create tight interconnections among investment firms, but is also likely to leave much of the industry reliant on two or three data models or data sets that would “herd” investment decisions in similar ways that would turn trickles of investment into floods with the potential to turn minor downturns into major disasters.

“This technology will be the center of future crises, future financial crises,” Gensler told Dealbook. “It has to do with this powerful set of economics around scale and networks.”