The effort stems from a broader review the agency conducted of the 2021 meme-stock frenzy. That episode raised alarms among regulators about the tactics investment platforms deployed to make the experience of trading feel more gamelike — using colorful graphics and other behavioral prompts to encourage more and riskier trades by retail investors that weren’t in their own best interest but generated fees for the financial middlemen.
Investment advisers already face a requirement to only recommend decisions in their clients’ best interests. The proposed rule goes further for online platforms, extending the ban on conflicts of interest to include features that use individuals’ data to try to direct their behavior.
“Artificial intelligence has complexity. But you have a basic, high-level strategic question: Are you optimizing just for investors, or are you optimizing also for the robo-advisor brokerage app?” SEC Chair Gary Gensler said. “That’s a straight-up conflict.”
The proposed rules would require investment firms to identify any potential conflicts of interest emerging from their use of AI and then eliminate them. And they would require firms to have written policies, procedures and record-keeping to prevent violations.
Steve Quirk, Robinhood’s chief brokerage officer, said the SEC’s proposal would make it harder for individuals to invest in stocks.
“The SEC’s proposal would turn back the clock, bringing U.S. financial markets to the old, manual days when retail investors were forced to interact with their broker or adviser by phone or at a branch office,” he said. “This isn’t in anybody’s best interest, least of all the new generation of retail investors.”
The agency’s two Republican commissioners also criticized the proposal, with one, Mark Uyeda, calling it “breathtakingly broad” and “wholly unnecessary.” The proposal’s “regulatory vagueness and considerable compliance challenges” may discourage innovation on Wall Street if it is approved, he said.
The public will have a 60-day window to weigh in before the five-member commission votes on a final version.
Gensler lately has been warning of the potential dangers that AI presents to financial stability. In a speech earlier this month, he said the tech could introduce new systemic risk, in part by promoting “herding” among investors who gather information from the same data and pile into trades that destabilize the market.
The promises and perils of AI have emerged as major concerns in Washington this year, but the policymaking on the matter has been fitful. The Federal Trade Commission has been among the first into the breach, launching an investigation earlier this month into OpenAI, examining whether the company behind the ChatGPT bot has violated consumer protection laws.