![Gary Gensler urges regulators to tame AI risks to financial stability Gary Gensler urges regulators to tame AI risks to financial stability](https://i0.wp.com/www.ft.com/__origami/service/image/v2/images/raw/https://www.ft.com/__origami/service/image/v2/images/raw/https%3A%2F%2Fd1e00ek4ebabms.cloudfront.net%2Fproduction%2F944887fd-714c-4815-875c-569b28d327eb.jpg?fit=%2C&ssl=1)
Stay informed with free updates
Just register AI myFT Digest – delivered straight to your inbox.
Regulators must quickly find a way to manage the risks to financial stability posed by the concentration of power in artificial intelligence platforms, the chairman of the U.S. Securities and Exchange Commission urged.
Gary Gensler told the Financial Times that without quick intervention, artificial intelligence will cause a financial crisis within a decade “almost inevitably”.
The SEC chairman said crafting artificial intelligence regulations will be a tough test for U.S. regulators because potential risks run through financial markets and stem from models developed by technology companies that don’t falls under the purview of Wall Street regulators.
“Frankly, it’s a tough challenge,” Gensler said. “This is a difficult financial stability problem to solve because most of our regulation is for individual institutions, individual banks, individual money market funds, individual brokers; that’s the nature of what we do. It’s a horizontal ( problem), many institutions may rely on the same underlying underlying model or underlying data aggregator.”
The U.S. Securities and Exchange Commission (SEC) proposed a rule in July to address potential conflicts of interest in predictive data analysis, but it focused on individual models deployed by broker-dealers and investment advisers.
Even if the current measures are updated, “this horizontal problem is still not solved”. . . What if everyone relies on a base model, and the base model is not at a broker-dealer but at one of the big tech companies,” Gensler said. “How many cloud providers are there in the country[that tend to offer AI as a service]? ”
He added: “I have raised this issue with the Financial Stability Board. I have raised this issue with the Financial Stability Oversight Council. I think this is indeed a cross-regulatory challenge”.
Regulators around the world are grappling with how to regulate artificial intelligence, as tech groups and their models are not naturally captured by specific regulators. The European Union has moved quickly to draft tough measures against the use of artificial intelligence in a groundbreaking law that is due to be fully approved by the end of this year. However, the U.S. is reviewing the technology to determine which aspects of it require new regulation and which aspects require compliance with existing laws.
Wall Street is already embracing artificial intelligence in a variety of ways, from robo-advice to the account-opening process and brokerage apps.
But Gensler worries that decisions made by all parties based on the same data model could lead to herd behavior, destabilizing finance and triggering the next crisis.
“I do think we’re going to have a financial crisis in the future . . . (and) after the action report people will say ‘Aha! Either there’s only one data aggregator or there’s only one model . . . we’ve been relying on’. Maybe It’s in the mortgage market. Maybe in some areas of the stock market,” Gensler said.
He added that AI’s powerful “network economics” make it “almost inevitable” and predicted that a crisis could occur as early as the late 2020s or early 2030s.
Lawmakers and regulators in Washington have stepped up scrutiny of artificial intelligence, raising concerns about market stability, data protection and antitrust. The Federal Trade Commission launched a review of ChatGPT maker OpenAI in July, focusing on consumer harm and data security. Antitrust agencies have warned that artificial intelligence’s structural dependence on scale could lead to technology monopolies.
Gensler, who works to address concentration issues in capital markets to improve efficiency, believes artificial intelligence could create competition problems in the space. “Could this lead to greater concentration of market makers?” he said.
The SEC also finalized a much-anticipated rule proposed in March 2022 that would require public companies to disclose their direct emissions and emissions from the energy they purchase, known as Scope 1 and Scope 2, respectively. Under the proposal, Scope 3 emissions, a broad measure that includes products companies buy from third parties, would only need to be reported if those products are deemed “significant” or part of a company’s climate goals.
Gensler declined to comment on whether Scope 3 disclosures will be included in the final version of the rule. Scope 3 disclosures were welcomed by investors but fiercely criticized by corporate America.
However, he said he would do the right thing for the American public by crafting a rule that is legal and court-backed based on economic conditions based on the comments the SEC has received. , based on trying to maintain a certain consistency with what has already happened.”
He noted that by 2021, 55% of companies in the Russell 1000 index had disclosed Scope 1 and 2 emissions.
The climate proposal has angered Republican lawmakers and attorneys general, with two dozen threatening to sue the SEC on the grounds that the SEC overstepped its authority, a claim Gensler denies. The chairman’s aggressive rulemaking agenda is facing other legal challenges, including a lawsuit filed by a coalition of private equity, venture capital and hedge fund groups seeking to block sweeping new measures targeting private equity managers.
Svlook