Beware, boasters: SEC challenges firms’ extravagant AI claims – Financial Planning

6 minutes, 20 seconds Read

The SEC is cracking down on a pair of advisory firms that made claims to be the first to bring AI-driven investing recommendations to the industry. The case, announced Monday as part of a $400,000 settlement, suggests how far the technology still has to go if it’s to be used for financial planning.

The Securities and Exchange Commission said on Monday that it had reached separate settlements with Global Predictions and Delphia over alleged false claims over their use of AI. Both firms were accused by the Wall Street watchdog of telling their clients they used artificial intelligence to guide investment decisions, when they were in fact relying on systems that differed little from older robo advisors.

San Francisco-based Global Predictions, for instance, advertised on its website that its proprietary technology could make “expert AI-driven forecasts.” The SEC said Global Predictions used an online chatbot that could communicate investment recommendations to clients but that the recommendations themselves came from robo advisors using conventional algorithms.

The SEC accused Toronto, Canada-based Delphia of making similar misstatements. The firm, according to regulators, collected client information for the ostensible purpose of feeding it into a large language model — a system capable of churning out convincing text.

“Delphia, however, never accomplished this goal,” according to the SEC. “While Delphia did collect certain client data intermittently between 2019 and 2023, it never used that data with artificial intelligence or machine learning or otherwise used that data in any way as inputs into its investing algorithms.”

Global Predictions and Delphia, which neither admitted to nor denied the SEC’s allegations, did not respond to requests for comment.

READ MORE:Don’t fear the robots: the case for deploying AI in wealth managementMorgan Stanley gives AI even bigger role with new appointmentRBC turns to AI to capture more of clients’ wallet shareCompliance experts look to AI, tech to help with WhatsApp violationsAI and 5 other FINRA worries for 2024

“We’ve seen time and again that when new technologies come along, they can create buzz from investors as well as false claims by those purporting to use those new technologies,” SEC Chairman Gary Gensler said in a statement.

This isn’t the first time Gensler has expressed concerns about exaggerated claims about “AI washing.” The term refers to firms that seek to exploit the current craze for artificial intelligence by exaggerating their reliance on the technology — much as some companies are accused of “green-washing” when they make equally unsubstantiated statements about how they’re benefiting the environment.

The SEC in July put forward a rule meant to make advisors explicitly responsible for eliminating conflicts of interest that might be buried in recommendations made by large language models, machine-learning systems and other types of AI. The proposal, which has come in for industry criticism, remains under consideration. 

Wally Okby, a strategic advisor in the the research and consulting firm Datos Insights’ wealth management practice, said he has seen very few firms saying they use AI for the purpose of making actual investment recommendations. When wealth managers make claims about the technology, it’s usually to say they’re relying on it to automate or accelerate certain internal features.

Morgan Stanley executives, for instance, have talked frequently about working with OpenAI’s GPT-4 large language model to help advisors pull answers to specific queries from the firm’s voluminous store of in-house research and analyst notes. 

“It tends to be functions that are low risk,” Okby said. “So it’s not coming out with advice on what to invest in, because that’s high risk. It’s things can be automated, so certainly manual tasks. It’s also ways to improve quality, reduce errors, accelerate processes and streamline the operating model.”

Okby said he believes the general public has a hard time distinguishing between what sorts of services can be provided by a true AI system and those that come from a standard robo advisor. That’s partly because, he said, the SEC and other regulators have yet to release a definition of what truly constitutes artificial intelligence.

“So is it machine learning, deep learning, neural networks, robotic personal assistance, virtual company chatbots, natural language processing?” Okby said.

Okby said that the distinction for him is that a robo advisor is truly just a machine that takes in information about clients — such as their net worth, expected retirement age and tolerance for risks — and uses an algorithm to automatically dispense investing recommendations. AI-driven advice, by contrast, will still require the participation of a person, he said.

Advisors, for instance, still have to check the accuracy and suitability of recommendations generated by an AI system. Artificial intelligence, he noted, remains notorious for its tendency to “hallucinate” — to make up seemingly factual statements. 

Okby said clients in turn should be skeptical of firms that claim to be drawing on the “power of AI” for recommendations.

“It’s still mainly a marketing buzzword used mainly to catch people’s attention,” he said. “In this case maybe it caught the wrong people’s attention.”

Besides making false claims about its use of AI, Global Predictions was alleged to have offered tax-loss harvesting services it in fact did not have. Tax-loss harvesting refers to the practice of using losses on stocks and other securities to offset capital gains taxes owed on investments that rose in value.

The SEC also alleged that Global Predictions, which was first federally registered in August 2023 and claimed to be the “first regulated AI financial advisor,” said on its website that it had $6 billion in assets under management. The SEC said the firm in fact reported in an annual regulatory filing that it had zero AUM.

The SEC further said Global Predictions made false claims about its investment recommendations in statements broadcast to a large audience through its website and YouTube. The firm, for instance, is alleged to have claimed its model offered a 3% to 6% “boost to returns.”

The SEC said many of these statements ran afoul of its marketing rule, which regulators began to enforce in November 2022. One provision of the rule prohibits firms from broadly releasing investment predictions that aren’t tailored to the goals and financial situations of individual clients.

The SEC also accused Global Predictions of making changes to its contracts with clients without giving proper notice and failing to follow procedures listed in its compliance manual, among other violations. To resolve the charges, the firm agreed to pay $175,000 and avoid future misstatements.

Delphia, whose registration with the SEC lapsed in January, was accused of similar violations. The firm reported in September that it was using its robo advisors to manage $7 million in 29,000 individual accounts and separately had $180 million in five pooled-investment vehicles. Those pooled investments have since been moved to another advisor, which the SEC didn’t identify.

In a press release from 2019, the firm claimed it was the ” “the first investment adviser to convert personal data into a renewable source of investable capital.” It also said “Delphia uses machine learning to analyze the collective data shared by its members to make intelligent investment decisions.”

The SEC said the firm acknowledged during an examination in 2021 that it had not in fact been using client data for the promised purposes. Despite that admission, according to the SEC, it continued making similar claims. Delphia agreed to settle the SEC’s charges for $225,000.

“As today’s enforcement actions make clear to the investment industry — if you claim to use AI in your investment processes, you need to ensure that your representations are not false or misleading,” Gurbir Grewal, the director of the SEC’s division of enforcement, said in a statement. “And public issuers making claims about their AI adoption must also remain vigilant about similar misstatements that may be material to individuals’ investing decisions.”

This post was originally published on this site

Similar Posts