Decoding the SEC’s First “AI-Washing” Enforcement Actions

Two recent SEC enforcement actions offer a first look at how the agency is approaching the use of artificial intelligence (“AI”) tools by registered firms. Against the backdrop of its pending proposed rules regarding predictive data analytics (“PDA”) and artificial intelligence, the SEC on March 18, 2024 announced settled charges against two investment advisers—Delphia (USA) Inc. (“Delphia”) and Global Predictions, Inc. (“Global Predictions”)—involving allegations that the firms’ promotional materials exaggerated their use of AI or machine learning in their investment services, a practice the SEC has described as “AI-washing.” Finding violations of the Investment Advisers Act rules governing marketing and compliance policies and procedures (Sections 206(2) and 206(4) and Rules 206(4)-1 and 206(4)-7 thereunder), the settled orders imposed civil penalties against Delphia and Global Predictions of $225,000 and $175,000, respectively.

The Delphia matter arose out of an examination by the SEC’s Division of Examinations. In the settled order, the SEC found that the firm falsely stated in its investor brochures, in a press release, and on its website that it deployed AI and machine learning to analyze client data through the algorithms used to manage client portfolios—but that no such inputs were actually used. For example, the order cited as false and misleading the statement that Delphia “uses machine learning to analyze the collective data shared by its members to make intelligent investment decisions,” alleging that current or prospective clients would find this a material consideration in utilizing an investment adviser. The order further charged that after its SEC exam, despite having informed the SEC that it would review and correct its disclosures, the firm made additional false and misleading representations in investor emails, social media posts, and a press release that it was using client data to inform its investment algorithms though it continued to lack such capabilities.

In the Global Predictions action, the SEC alleged, among other things, that the firm’s website, social media, and client emails falsely described it as the “first regulated AI financial advisor.” The settled order also found that the adviser promoted the use of “AI-driven forecasts” to inform investment recommendations—none of which it in fact employed. As part of the settlement, which involved these and other unrelated violations, Global Predictions agreed to retain a compliance consultant to review its marketing and training materials.

The settlements call to mind SEC Chair Gensler’s promise earlier this year to crack down on “AI washing.” In recent years, the SEC has warned of the dangers of corporate “washing,” a practice where corporate boards allegedly falsely promote metrics on various issues (such as environmental social and governance claims). This pair of enforcement actions portend that the SEC may pursue claims of “AI-washing” in a fashion similar to its pursuit of alleged “ESG greenwashing.”

The orders further echo concerns that appear to inform the SEC’s proposed PDA conflicts of interest rules. If enacted, these pending amendments (to Sections 15 and 17 of the Securities Exchange Act and Sections 204 and 211 of the Advisers Act) would impose significant new requirements on the use of PDA technologies by broker dealers and investment advisers, reflecting the SEC’s stated view that such technologies could promote investments that advance firms’ interests over the interests of investors or clients. Dissenting SEC Commissioner Hester Peirce has cautioned that the proposed amendments would encompass even “spreadsheets, commonly used software, math formulas, statistical tools, and AI trained on all manner of datasets.”

The Delphia and Global Predictions settlements do not directly implicate the reported concerns underlying the proposed PDA rule amendments and are instead based on existing statutes, rules governing advisers’ fiduciary duties, and marketing and compliance obligations. But animating both the proposed rules and these first two enforcement actions appears to be the common theme that the SEC considers how firms use AI technologies—and whether their uses of such tools are fully understood and accurately portrayed—to be important for current or prospective investors and clients.

As the first shot across the bow, the settlements underscore the SEC’s commitment to using existing rules to regulate firms’ various uses of AI tools, particularly in the area of “AI-washing.” This is further illustrated by the AI sweep reportedly commenced out of the SEC’s Division of Examinations in December 2023 (which seeks detailed information about firms’ use of AI). Investment advisers should not only ensure that their marketing and other materials convey information about AI that is consistent with their actual use of such technologies but should also maintain documentation of their support for any such claims. As with its stance on ESG greenwashing, the SEC will likely consider any and all disclosures made regarding AI—whether in SEC filings, client communications, social media, or on websites—when evaluating potential AI-washing claims. Advisers and investment firms need to coordinate internally to ensure the consistent description of any AI tools in a manner aligned with their actual use by investment professionals.