SEC Chair Outlines AI Vision in First Dedicated Address

In his first extended public remarks devoted to artificial intelligence (AI) since taking the helm of the agency nearly one year ago, SEC Chair Paul Atkins on March 4 outlined efforts to more fully embed AI technology across the agency, declaring that he seeks to “ingrain innovation into the SEC’s culture, broadly and deliberately.” Atkins was speaking before an AI innovation series roundtable on strategy and governance principles convened by the Financial Stability Oversight Council (FSOC).

Atkins discussed the mission and objectives of the SEC’s AI Task Force, how the agency is integrating AI technology into its enforcement program and its approach to AI-related disclosure.

“The main message that I want to leave with you today,” he said, “is that AI is more than an instrument of efficiency or convenience. It is a force that stands to enable investors to participate in the markets with greater confidence, businesses to allocate capital with sharper precision, and regulators to oversee those financial markets with deeper insight.”

Noting that the SEC has grappled with the complexities of technological advances in the past, Atkins emphasized that financial regulators must keep pace with the markets they oversee: “We must strive to keep up, and to collapse the distance between the regulators and the regulated.” Accordingly, “for our part today,” he declared, “we are not content to retreat from the AI revolution, nor to remain tethered to the tools of a bygone era. Instead, our posture at the SEC is clear: we intend to understand AI; to assess its potential; and, where appropriate, to adopt its solutions.”

SEC’s AI Task Force: Mission and Objectives 

To those ends, the SEC established its AI Task Force last August “to facilitate the development and deployment of AI across the Commission,” centralizing efforts to accelerate AI integration and foster cross-disciplinary collaboration. The Task Force’s mandate includes tools to:

• • conduct risk assessments for potential examination;

• • detect potential market misconduct, such as fraud and rule violations;

• • review disclosures with greater speed and efficiency;

• • react to public input on new proposals; and

• • evaluate market-wide risks that bear on U.S. capital markets.

Atkins stressed that these AI-powered tools must be deployed responsibly in the SEC’s enforcement program, with appropriate safeguards for due process and human judgment.

Enforcement: ‘Misconduct Remains Misconduct’

Atkins underscored that the SEC is “committed to using AI-enabled tools and systems in ways that augment our work responsibly. Human interaction is still required, indeed imperative, at every stage of our risk assessment program. Due process demands it,” he said, emphasizing that algorithmic tools cannot supplant the considered judgment of commissioners and staff or serve as the sole basis for an SEC enforcement action.

Atkins also lamented the temptation of abuse that invariably accompanies advances in technology, observing that “[b]ad actors have begun to exploit AI and the buzz that surrounds it.” He pledged that the SEC will seek to hold accountable those that misuse AI technologies to further fraudulent and manipulative schemes and noted the SEC has already brought actions against bad actors for deception that involves false, misleading or exaggerated claims about the use of AI in their products and services. “In short, while the mechanisms of fraud may change, our obligation does not,” he said. “The Commission’s mandate to protect investors is technology neutral. And misconduct remains misconduct, regardless of the medium.”

Disclosure: ‘Materiality-Based Transparency’

Reinforcing that the SEC brings the same measured regulatory philosophy to both enforcement and disclosure, Atkins stated that principles-based (versus prescriptive) rules—rooted in materiality—have historically served as the SEC’s optimal approach, reflecting his broader stated vision for the SEC to provide the “minimum effective dose of regulation” needed to elicit the information that is material to investors. He added that this time-tested approach should guide how public companies disclose AI-related developments, just as it guides disclosures about any other development. “The standard is a familiar one: whether there is a substantial likelihood that a reasonable shareholder would consider the information important in making an investment decision.” These remarks align with Atkins’s repeated emphasis that financial materiality should be the guiding principle or “north star” of the SEC’s disclosure regime.

“Prescriptive mandates are not the answer to every emerging technology,” he continued. “And disclosure ‘checklists’ are no substitute for materiality-based transparency that offers meaningful disclosure under established principles. If the advent of each new technology becomes a pretext for new line items, then disclosure swiftly loses its discipline. In the absence of a limiting principle, a morass of information can do more to obscure than to illuminate.”

Atkins’s remarks reinforced the position he articulated last December at the SEC’s Investor Advisory Committee meeting when the committee voted to approve a nonbinding recommendation urging the Commission to standardize disclosures regarding the impact of AI on public company operations. Citing the lack of comprehensive guidance on this topic and “uneven and inconsistent” reporting across industries, the committee recommended that the SEC integrate AI disclosure guidance into the existing Regulation S-K framework that would require companies to:

• • Define terms. Clearly define what they mean when they use the term “Artificial Intelligence” in their disclosures;

• • Disclose oversight. Disclose board mechanisms, if any, for overseeing AI deployment at the company; and

• • Report impact. Separately report, if material, on AI deployment and its effects on both internal business operations and consumer-facing matters.

At the time, Atkins and two other commissioners each signaled skepticism toward the recommendation, maintaining that the SEC’s existing principles-based rules are sufficient. Atkins argued against creating prescriptive disclosure requirements for every “new thing,” stating that current rules already allow companies to inform investors of material risks associated with new developments—including how AI affects their financial results, how AI can be a material risk factor to an investment and how AI is a material aspect of their business model—without “ever-expanding checklists.” Commissioners Peirce and Uyeda voiced similar concerns, with Peirce questioning why the SEC’s disclosure regime should force conformity when AI adoption varies across companies and industries, and Uyeda cautioning against using the federal securities laws as “a backdoor attempt to regulate AI.”

Although the current Commission is unlikely to act on the recommendation, it offers a useful framework for companies developing AI-related disclosure and a preview of the kinds of questions sophisticated investors may ask.

Atkins concluded by encouraging market participants to engage with Commission staff around innovative AI use cases and to share input on how technological advances can further the agency’s three-part mission to protect investors; maintain fair, orderly and efficient markets; and facilitate capital formation.