SEC Finalizes AI Disclosure Requirements for Investment Advisers and Broker-Dealers

SEC Finalizes AI Disclosure Requirements for Investment Advisers and Broker-Dealers

The Securities and Exchange Commission voted 3-2 today to adopt final rules requiring investment advisers and broker-dealers to disclose their use of artificial intelligence systems, marking the most comprehensive AI regulation in financial services to date. The rules, which take effect July 1, 2025, mandate detailed disclosures about AI usage in investment advice, trading, client communications, and compliance functions.

Key Requirements

The final rules establish several core disclosure obligations:

AI System Inventory: Firms must maintain a current inventory of all AI systems used in client-facing or investment-related activities. The inventory must document each system's purpose, data sources, decision-making authority, and human oversight mechanisms.

Client Disclosures: When AI systems contribute to investment recommendations or portfolio management decisions, firms must disclose this fact to clients in plain English. Disclosures must explain the AI system's role, limitations, and how human advisers supervise or override automated recommendations.

Conflicts of Interest: Firms must identify and disclose conflicts of interest arising from AI system design or implementation. This includes scenarios where AI optimization objectives may not align perfectly with client interests, such as systems optimized for firm profitability alongside client returns.

Risk Management Framework: Firms must establish written policies and procedures governing AI system development, testing, deployment, monitoring, and deactivation. Annual reviews of AI risk management frameworks are required, with board-level oversight for larger firms.

Compliance Timeline

The SEC structured implementation in phases:

• July 1, 2025: Initial AI system inventory due for existing systems
• October 1, 2025: Client disclosure requirements take effect
• January 1, 2026: Full risk management framework documentation required
• April 1, 2026: First annual AI risk management review due

"The phased approach recognizes that many firms are still developing their AI governance capabilities," explained SEC Chair Gary Gensler. "We want to ensure compliance is thoughtful and effective, not rushed."

Industry Reaction

Industry responses have been mixed. Larger firms with established AI governance programs generally support the rules, while smaller advisers express concern about compliance costs.

"These rules provide helpful clarity," said Thomas Anderson, Chief Compliance Officer at a major investment bank. "We've been building AI governance frameworks proactively, and now we have regulatory certainty about expectations."

Smaller registered investment advisers voiced different concerns. "The compliance burden is significant for firms using even basic AI tools like automated portfolio rebalancing," noted Patricia Collins, owner of a boutique RIA. "We may need to hire additional compliance staff or invest in specialized technology."

Technical Challenges

Several technical aspects of the rules have generated discussion:

AI Definition Scope: The rules define AI broadly to include machine learning, natural language processing, and traditional algorithmic systems with adaptive capabilities. Some firms questioned whether this captures routine automation that predates modern AI.

Third-Party Systems: When using vendor-provided AI systems, firms remain responsible for required disclosures even if they lack full visibility into system internals. This has prompted calls for enhanced vendor due diligence requirements.

Testing and Validation: The rules require periodic testing and validation of AI systems but provide limited guidance on acceptable methodologies. Industry groups are developing best practice frameworks to fill this gap.

Global Context

The SEC's rules align with emerging international approaches to AI regulation in financial services:

The UK's FCA published similar guidance in 2024, emphasizing algorithmic accountability and consumer protection. The European Union's AI Act, which took effect in December 2024, classifies certain financial AI systems as "high-risk," triggering stringent requirements.

Singapore's MAS issued AI governance principles in early 2024, focusing on fairness, ethics, accountability, and transparency—the FEAT framework. The SEC rules incorporate similar principles while adding specific disclosure mandates.

"There's clear international convergence on AI governance in financial services," observed Dr. James Liu, a financial regulation expert at Columbia Law School. "Firms operating globally face overlapping requirements, making comprehensive AI governance essential."

Enforcement Approach

The SEC indicated it would take a principles-based enforcement approach initially, focusing on good-faith compliance efforts rather than technical violations during the first year. However, Chair Gensler emphasized that firms demonstrating inadequate AI risk management or misleading client disclosures would face enforcement action.

Recent SEC examinations have already identified AI-related deficiencies at several firms. Examiners found instances where firms marketed AI capabilities that didn't exist, used AI systems without appropriate testing, and failed to supervise AI-generated client communications.

Technology Solutions

The new requirements are driving demand for specialized compliance technology. RegTech vendors including RuleWise are developing solutions to help firms maintain AI system inventories, generate required disclosures, and document risk management frameworks.

"The rules create both challenges and opportunities," said Emma Thompson of RuleWise. "Firms need technology to manage compliance with AI rules while simultaneously using AI to improve their overall compliance operations. We're helping clients navigate that complexity."

Looking Ahead

The SEC's AI rules represent the first comprehensive federal regulation of artificial intelligence in financial services, but likely not the last. Congressional committees are considering broader AI legislation, and other financial regulators including the CFTC and Federal Reserve are developing their own frameworks.

For investment advisers and broker-dealers, the message is clear: AI governance is no longer optional. Firms must implement robust frameworks for managing AI risks, ensuring transparency with clients, and maintaining regulatory compliance as AI adoption accelerates across the industry.

The full text of the final rules and accompanying guidance documents are available on the SEC website, with a compliance checklist and FAQ document published to assist firms with implementation.