Key Takeaways:
- On March 18, 2024, the U.S. Securities and Exchange Commission (“SEC”) announced settled charges against two investment advisers, Delphia (USA) Inc. (“Delphia”) and Global Predictions Inc. (“Global Predictions”) for making false and misleading statements about their alleged use of artificial intelligence (“AI”) in connection with providing investment advice.
- These settlements are the SEC’s first-ever cases charging violations of the antifraud provisions of the federal securities laws in connection with AI disclosures, and also include the first settled charges involving AI in connection with the Marketing and Compliance Rules under the Investment Advisers Act of 1940 (“Advisers Act”).
- The matters reflect Chair Gensler’s determination to target “AI washing”—securities fraud in connection with AI disclosures under existing provisions of the federal securities laws—and underscore that public companies, investment advisers and broker-dealers will face rapidly increasing scrutiny from the SEC in connection with their AI disclosures, policies and procedures.
- We have previously discussed Chair Gensler’s scrutiny of AI washing and AI disclosure risk in Form ADV Part 2A filings. In this client alert, we discuss the charges and AI disclosure and compliance takeaways.
On March 18, 2024, the U.S. Securities and Exchange Commission (“SEC”) announced settled charges against two investment advisers, Delphia (USA) Inc. (“Delphia”) and Global Predictions Inc. (“Global Predictions”) for making false and misleading statements about their alleged use of artificial intelligence (“AI”) in connection with providing investment advice. These settlements are the SEC’s first-ever cases charging violations of the antifraud provisions of the federal securities laws in connection with AI disclosures, and also include the first settled charges involving AI in connection with the Marketing and Compliance Rules under the Investment Advisers Act of 1940 (“Advisers Act”). The matters reflect Chair Gensler’s determination to target “AI washing”—securities fraud in connection with AI disclosures under existing provisions of the federal securities laws—and underscore that public companies, investment advisers and broker-dealers will face rapidly increasing scrutiny from the SEC in connection with their AI disclosures, policies and procedures. We have previously discussed Chair Gensler’s scrutiny of AI washing and AI disclosure risk in Form ADV Part 2A filings. In this client alert, we discuss the charges and AI disclosure and compliance takeaways.
THE SEC SETTLEMENTS
The SEC charged both Delphia and Global Predictions with violations of Section 206(2) of the Advisers Act—an antifraud provision—for false and misleading claims about the use of AI in connection with their investment advisory services, as well as with Marketing Rule and Compliance Rule failures in connection with these misstatements.
Antifraud Charges
The SEC found that Delphia claimed in its Form ADV Part 2A, in a press release and on its website that it used AI and machine learning to analyze its retail clients’ spending and social media data to inform its investment advice when it actually did not use any such data in its investment process. Delphia’s false statements included declarations that it used “a predictive algorithmic model” for asset selection and deployed “machine learning to analyze the collective data shared by its members to make intelligent investment decisions.” Notably, even after the Division of Examinations discovered the misleading statements in a prior exam and Delphia admitted that it had not ever created an algorithm to use client data for investment decisions, it nonetheless did not fully comply with required Form ADV brochure updates about these disclosures and continued to make similar false and misleading statements in client communications and marketing materials about its nonexistent “proprietary algorithms” and predictive data analytics for almost two years thereafter.
The SEC found that Global Predictions made false and misleading claims about its use of AI and the services that it offered. Global Predictions had claimed on its website that its technology incorporated “[e]xpert AI-driven forecasts,” when it did not, and inaccurately claimed to be the “first regulated AI financial advisor” on its website, in emails to current and prospective clients, and on social media.
Compliance Rule and Marketing Rule Violations
The SEC charged Delphia for violations of Section 206(4) of the Advisers Act and Rules 206(4)-7 thereunder (the “Compliance Rule”) and 206(4)-1 (the “Marketing Rule”) thereunder for its failures to implement policies and procedures necessary to ensure that the advertisements that it published, circulated or distributed were accurate and did not contain misleading or false statements, particularly on social media. Delphia also failed to establish a clear advertising review and approval process, which led to its continued AI-related misrepresentations even after the examination findings. The SEC likewise charged Global Predictions under the Compliance Rule and the Marketing Rule for failing, among other things, to substantiate claims on its website that it was the “first regulated AI financial advisor” and to implement its policies and procedures requiring the review and approval of all marketing material.
TAKEAWAYS FOR MITIGATING SEC ENFORCEMENT RISK IN CONNECTION WITH AI DISCLOSURES
While the SEC’s proposed predictive data analytics rule has generated considerable debate, the settlements demonstrate that the SEC does not need to wait for an AI-specific rule to charge AI-related disclosure and related violations, but can readily deploy the existing federal securities laws to do so. The settlements also reinforce that AI is a priority not only for the Division of Examinations but also for the Division of Enforcement.
Public companies, investment advisers and broker-dealers accordingly may want to consider the following best practices for disclosure and governance to avoid AI-related exam deficiencies and to best position themselves in the event of an enforcement investigation by an increasingly aggressive SEC:
AI Disclosure Best Practices
- Be Clear and Accurate as to What You Do (And Don’t) Use AI For. AI usage varies widely among market participants as a function of each firm’s business model. There is no “one-size-fits all” AI disclosure and companies and firms must be able to accurately articulate their AI use cases—and avoid understatement or overstatement. Descriptions of AI use should be consistent across all media, corporate and marketing communications and regulatory filings.
- Avoid Using Hypothetical Language for Actual AI Practices. Using hypothetical language to indicate the possibility of a certain AI use case can give rise to both examination and enforcement scrutiny. Companies and firms should avoid using hypothetical or qualifying language like “may” to describe AI use cases that actually exist. The SEC has brought numerous cases against companies and firms for using hypothetical language to describe actual practices, and these enforcement actions will serve as a template for future SEC inquiries involving AI practices. Given the SEC’s enhanced scrutiny of AI disclosures, market participants should carefully consider whether to include such disclosures and, if so, how to frame them to avoid claims that those disclosures are misleading. In addition, a company or firm that uses such hedging language cannot “set-it-and-forget-it,” and should consider updating such language in future filings if AI use transforms from theoretical to actual.
- Understand and Accurately Disclose the Risks Associated with AI Use: As more companies and firms adopt AI (including generative AI) as part of their core business functions, certain well-known risks associated with generative AI persist, such as quality control, privacy, IP, data-use limitations, cybersecurity, bias and transparency. Accordingly, disclosures should be clear, comprehensive and precise about such risks. As seen in matters involving the improper use of hypothetical language, the SEC has charged firms for using hypothetical language to describe risks that have materialized. Companies and firms should similarly exercise caution and accuracy in disclosing AI risks that have emerged.
AI Governance Best Practices
- Define AI Internally, Including for Marketing Purposes. Consider creating an internal definition of AI to avoid claims of misrepresenting AI usage. A definition of AI can also limit the risk that personnel will inadvertently mischaracterize simple automated systems as AI in advertising or marketing materials.
- Training. Provide training for internal marketing and business development personnel on the legal and reputational risks of overselling AI capabilities. Because AI disclosure risk has rapidly become an important compliance issue, educating marketing and compliance teams on the regulatory scrutiny will benefit risk management.
- Review. Consider establishing a process for reviewing all current and proposed disclosures about AI to ensure that they are accurate, can be substantiated and do not exaggerate or overpromise. The review process may require both a legal and a technical review. Data and AI counsel may wish to take a more active role in marketing review as this process develops.
- Vendor Considerations. For AI systems that are provided by a vendor, be careful not to merely repeat vendor claims about the AI system without taking appropriate steps to verify each claim’s accuracy.
- Risk Assessments. For high-risk AI systems, consider conducting impact assessments to determine foreseeable risks and how best to mitigate those risks, and then consider disclosing those risks in external statements about the AI systems.
This publication is for general information purposes only. It is not intended to provide, nor is it to be used as, a substitute for legal advice. In some jurisdictions it may be considered attorney advertising.