Key Takeaways
- • The SEC has withdrawn its "predictive data analytics" proposal but intensified focus on AI oversight, shifting emphasis from conflicts of interest (duty of loyalty) to fiduciary duty of care—requiring advisers to demonstrate they understood, tested, and supervised AI tools to ensure client-first decision-making.
- • The SEC Division of Investment Management and Division of Examinations are prioritizing examination of registrants' use of automated investment tools, AI technologies, and trading algorithms in 2026, with particular scrutiny on transparency, auditability, and consistency with fiduciary duties.
- • Advisers must establish explainability frameworks ensuring investment personnel can articulate in plain terms what AI tools do, their material limitations, and how outputs are weighted—avoiding "black box" treatment that masks model logic from human oversight.
- • Documentation of intended use cases, material features, and change management procedures is critical to prevent AI models from being repurposed beyond original scope, where regulatory and operational failures commonly emerge.
- • Advisers face heightened compliance obligations regarding training data (particularly the exclusion of material non-public information), model opacity, governance structures, and data security under Regulation S-P amendments—with examination risk concentrated on firms unable to trace how customer information flows through and is transformed by AI systems.
Summary
Registered investment advisers are rapidly integrating artificial intelligence into portfolio construction, trading, research, screening, and client communications, but regulatory scrutiny is intensifying despite the SEC's withdrawal of its proposed "predictive data analytics" rule. The withdrawal does not signal reduced supervisory attention; rather, the SEC is recalibrating its enforcement lens from conflicts of interest toward fiduciary duty of care—a fundamental shift that will require advisers to demonstrate rigorous understanding, testing, and supervision of AI tools to ensure decisions remain aligned with client interests.
The SEC Division of Investment Management has elevated AI to a central supervisory priority. In December remarks, the Division Director characterized AI as a "transformative force" warranting thoughtful regulation and support. His subsequent January remarks on proxy voting highlighted the possibility that advisers may deploy AI in voting processes, provided such use incorporates transparency, auditability, and fiduciary consistency. The Division of Examinations' 2026 priorities explicitly target registrants' use of automated investment tools, AI technologies, and trading algorithms, signaling active compliance examination activity throughout the year.
The regulatory framework does not yet contain AI-specific rules, but the core fiduciary principles remain unchanged: advisers must ensure advice serves clients' best interests, enforce policies ensuring personnel compliance with legal standards, and provide sufficiently clear disclosures for informed investor consent. However, AI introduces two material complications: speed and scale of decision-making, and opacity of implementation. To address these risks, advisers should prioritize five operational and compliance considerations. First, explainability requires investment personnel to articulate—in plain terms—what tools are designed to accomplish, what information they rely on, material limitations, and how outputs are weighted against other analysis. Personnel need not possess computer science expertise but must understand which data feeds the model, recognize when it is underperforming, and know when outputs should be overridden. Second, documentation of intended use cases, material features, and change management procedures must be preserved, particularly for internally developed tools and significant modifications to existing models—this creates an audit trail protecting against scope creep and unintended repurposing. Third, model transparency and validation are essential; advisers must understand patterns the model detects, training data composition, and historical failure scenarios. Critical compliance risk centers on whether material non-public information has been inadvertently or intentionally ingested, creating potential violations of Section 204A of the Advisers Act. Fourth, governance frameworks should designate clear authority for AI tool deployment and establish which AI outputs constitute books and records under Rule 204-2. Finally, privacy and data security require advisers to map how customer information flows into, through, and out of AI systems, particularly given the SEC's recent Regulation S-P amendments and the Division of Examinations' stated intention to review information security and data privacy during compliance examinations. The opacity of many AI tools complicates this assessment, particularly when firms grant models full access to all adviser records.
The SEC Division of Investment Management has elevated AI to a central supervisory priority. In December remarks, the Division Director characterized AI as a "transformative force" warranting thoughtful regulation and support. His subsequent January remarks on proxy voting highlighted the possibility that advisers may deploy AI in voting processes, provided such use incorporates transparency, auditability, and fiduciary consistency. The Division of Examinations' 2026 priorities explicitly target registrants' use of automated investment tools, AI technologies, and trading algorithms, signaling active compliance examination activity throughout the year.
The regulatory framework does not yet contain AI-specific rules, but the core fiduciary principles remain unchanged: advisers must ensure advice serves clients' best interests, enforce policies ensuring personnel compliance with legal standards, and provide sufficiently clear disclosures for informed investor consent. However, AI introduces two material complications: speed and scale of decision-making, and opacity of implementation. To address these risks, advisers should prioritize five operational and compliance considerations. First, explainability requires investment personnel to articulate—in plain terms—what tools are designed to accomplish, what information they rely on, material limitations, and how outputs are weighted against other analysis. Personnel need not possess computer science expertise but must understand which data feeds the model, recognize when it is underperforming, and know when outputs should be overridden. Second, documentation of intended use cases, material features, and change management procedures must be preserved, particularly for internally developed tools and significant modifications to existing models—this creates an audit trail protecting against scope creep and unintended repurposing. Third, model transparency and validation are essential; advisers must understand patterns the model detects, training data composition, and historical failure scenarios. Critical compliance risk centers on whether material non-public information has been inadvertently or intentionally ingested, creating potential violations of Section 204A of the Advisers Act. Fourth, governance frameworks should designate clear authority for AI tool deployment and establish which AI outputs constitute books and records under Rule 204-2. Finally, privacy and data security require advisers to map how customer information flows into, through, and out of AI systems, particularly given the SEC's recent Regulation S-P amendments and the Division of Examinations' stated intention to review information security and data privacy during compliance examinations. The opacity of many AI tools complicates this assessment, particularly when firms grant models full access to all adviser records.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.