SEC Targets “AI Washing” by Companies, Investment Advisers, and Broker-Dealers

What Is “AI Washing”?

With consumer and investor interest in artificial intelligence (AI) surging, there has been heightened concern among regulatory agencies, including the Federal Trade Commission (FTC) and the Securities and Exchange Commission (SEC), with companies engaging in a deceptive practice the FTC and SEC have dubbed “AI washing.” Specifically, as used by the SEC in recent enforcement actions and public statements, AI washing refers to companies, investment advisers and broker-dealers’ making false, misleading, or exaggerated claims about the AI-related capabilities of their products and services. Like “greenwashing,” where companies make false or exaggerated claims about the sustainability benefits of their products and services to exploit the growing base of environmentally conscious consumers and investors, companies engaging in AI washing seek to exploit increasing consumer and investor interest in AI.

Recent SEC Enforcement Actions

On March 18, 2024, the SEC announced settled charges against two investment advisers, Delphia (USA) Inc. (Delphia) and Global Predictions Inc. (Global Predictions), which included making false and misleading statements about their use of AI. Without admitting or denying the charges, Delphia and Global Predictions agreed to pay civil penalties of $225,000 and $175,000, respectively, to settle with the SEC.

The SEC charged Delphia, a former SEC-registered investment adviser, with making false and misleading statements spanning a three-year period about using AI and machine learning to collect client data (including from social media, banking, credit card, and online purchases) and using that client data as inputs into its investment algorithms. The SEC found that while Delphia collected such client data intermittently between 2019 and 2023, it never used that client data with artificial intelligence or machine learning or otherwise used client data as inputs into its investing algorithms. Delphia’s false and misleading statements quoted by the SEC in its settlement offer and cease-and-desist order include a claim in a press release that Delphia “uses machine learning to analyze collective data shared by its members to make intelligent investment decisions”; a statement on its website that Delphia “put[s] collective data to work to make [its] artificial intelligence smarter so it can predict which companies and trends are about to make it big and invest in them before everyone else”; and a social media post stating that Delphia’s “proprietary algorithm uses the data being invested by [its] members, so [it] can make stock selections across thousands of publicly traded companies up to seven financial quarters in the future.” The SEC found Delphia’s misleading statements to be material because Delphia represented to current and prospective investors that its use of client data as inputs in its investing algorithms was a key characteristic distinguishing it from other advisers. Additionally, Delphia failed to adopt and implement policies necessary to ensure its advertisements were accurate and did not contain misleading or untrue statements.

Similarly, the SEC brought an enforcement action against Global Predictions, a San Francisco–based investment adviser, for making false and misleading statements—including regarding its use of AI—on its website, on social media sites, and in emails to current and prospective clients, which constituted advertisements. Specifically, according to the SEC’s settlement offer and cease-and-desist order, Global Predictions falsely stated on its website that its technology incorporated “[e]xpert AI-driven forecasts.” Additionally, the SEC found that Global Predictions inaccurately claimed to be the “first regulated AI financial advisor” on its website, on various social media sites, and in emails to current and prospective clients and that Global Predictions could not produce documents to substantiate this claim.

The SEC’s Public Statements

Along with the settlements with Delphia and Global Predictions, the SEC and its officials have made several public statements regarding AI washing in recent months.

On January 25, 2024, the SEC Office of Investor Education and Advocacy, the North American Securities Administrators Association, and the Financial Industry Regulatory Authority issued a joint investor alert to raise awareness of the increase in investment frauds involving the purported use of AI and other emerging technologies.

On February 2, 2024, the SEC settled a fraud case against Brian Sewell and his company, Rockwell Capital Management, for raising $1,200,000 for a fund to invest in cryptocurrency trading that was purported to employ AI and machine-learning technology that never existed. Without admitting or denying the charges, Rockwell Capital Management agreed to pay disgorgement and prejudgment interest totaling $1,602,089, and Mr. Sewell agreed to pay a civil penalty of $223,229. In a press release announcing the settlement, Gurbir S. Grewal, Director of the SEC’s Division of Enforcement, stated that “the SEC will continue to hold accountable those who claim to use attention-grabbing technologies to attract and defraud investors.”

On February 13, 2024, in a speech at Yale Law School, SEC Chair Gary Gensler emphasized that companies must be truthful about their use of AI and that “claims about prospects should have a reasonable basis and investors should be told that basis.” He also cautioned against using boilerplate language in AI-related risk disclosures, as material risks about AI tend to be particularized to a company. He also noted the possible need for companies to define for investors what they mean when referring to “AI” in their public statements to prevent misunderstanding. Lastly, he explicitly warned companies that misleading the public by making false statements about their AI models can violate securities laws.

On March 18, 2024, the same day the SEC announced its settlements with Delphia and Global Predictions, Mr. Gensler stated in a video that “AI washing may violate the securities laws” and emphasized that public companies, investment advisers, and broker-dealers should “say what they’re doing and do what they’re saying” regarding their use of AI. He warned investment advisers and broker-dealers not to mislead the public by claiming they are using an AI model when they are not, or by saying they are using an AI model in a particular way but failing to do so. He advised public companies to make sure they have a reasonable basis for the claims and particular risks they face regarding their AI use, and that investors should be informed of that basis. On the same date, in a video announcing the charges against Delphia and Global Predictions, Mr. Grewal echoed Mr. Gensler’s message on AI washing. He acknowledged and cautioned investors that an investment adviser’s claims about its use of AI may precede its actual development or implementation of AI. Mr. Grewal also emphasized that the enforcement actions against Delphia and Global Predictions should serve as notice to the investment industry that its representations regarding AI should not be false or misleading.

Key Takeaways

Regulatory bodies, including the SEC, are in the early stages of regulating the use of AI, and the nature of enforcement cases and the extent of penalties remain uncertain. However, the SEC will likely continue to bring enforcement actions related to AI washing as consumer and investor interest in AI grows. To ensure accurate and complete AI-related disclosures and public statements, companies should strive to do the following:

1. Clearly differentiate current uses of AI from future plans to use AI. If a company discusses plans to develop or use AI, it should use reasonable assumptions and disclose such assumptions to investors.
2. Only promote AI capabilities that can be substantiated and avoid exaggerations. If a statement can be substantiated, keep a record of the supporting documents or materials.
3. Be specific about how and where AI is being used and whether AI-related capabilities are being developed in house or supplied by third parties.
4. Understand how the public defines and understands the term “AI” to avoid misleading statements. If a company uses the term “AI” in a way that differs from popular understanding, it should clarify what it means by “AI” to prevent confusion.
5. Identify potential and existing risks associated with the use of AI and be clear and comprehensive when making disclosures. While the SEC limited its findings to affirmative misrepresentations in its enforcement actions against Delphia and Global Predictions, future investigations could relate to material omissions by a company about the risks associated with its use of AI.
6. Avoid using boilerplate language when disclosing AI-related risks, as such risks tend to be company-specific.
7. Review company websites, social media accounts, marketing materials, regulatory filings, press releases, and public statements to confirm that AI-related statements and disclosures are accurate and up to date.
8. Train marketing personnel to identify and avoid AI washing and educate them about the associated risks.
9. Establish internal review procedures to ensure that future AI-related public statements are accurate and consistent across company websites, social media accounts, marketing materials, regulatory filings, press releases, and public statements.