The SEC Withdrew Its AI Conflicts Rule. Here's What Still Applies to Advisers

By The Professions Cloud Editors Updated October 2026 · 7 sources · 3 min read

Prices and facts checked October 13, 2026. We don't accept payment for rankings.

What happened: on June 12, 2025, the SEC formally withdrew 14 pending rule proposals from the previous administration, including the "Conflicts of Interest Associated with the Use of Predictive Data Analytics" proposal that would have governed advisers' use of AI (Paul Hastings; National Law Review).

Why it still matters today: the withdrawal removed a new AI rulebook, not the existing one. The SEC kept bringing "AI washing" cases, set up a dedicated unit for emerging-technology fraud, and made AI a named exam priority for 2026. If you use AI tools or talk about AI in your marketing, the rules below still apply.

What the withdrawn rule would have done

The SEC proposed the rule on August 9, 2023 (National Law Review). It would have required broker-dealers and advisers to identify conflicts of interest created by predictive analytics (including AI) in investor interactions, and to "eliminate or neutralize" them rather than disclose them (Wealth Management, Jun 13, 2025).

Industry pushback was heavy. The Investment Adviser Association warned it would create "a brand new framework for handling conflicts" that could undercut advisers' fiduciary duty, and smaller firms worried about the compliance burden (Wealth Management).

Other adviser proposals withdrawn the same day included safeguarding (custody), cybersecurity risk management and third-party outsourcing (Paul Hastings). Any future SEC rule on these topics has to start over with a new proposal and comment period.

What still applies: four things

1. "AI washing" is still enforced

The SEC's first AI-washing cases came in March 2024: Delphia paid $225,000 and Global Predictions $175,000 for claims about AI they couldn't back up, under the antifraud provisions and the Marketing Rule (Morrison Foerster). A later settled case against a small adviser, its parent, CEO and a board member involved marketing an AI trading platform the adviser didn't have. The CEO paid over $460,000 and took a five-year industry bar (Seward & Kissel).

2. A dedicated enforcement unit

On February 20, 2025, the SEC created the Cyber and Emerging Technologies Unit, about 30 specialists whose remit explicitly includes fraud committed using AI and machine learning (SEC).

3. Exams

AI is on the Division of Examinations' list for 2026, including whether firms' representations about AI are accurate and whether policies cover AI use (Foley). See our breakdown of the 2026 priorities.

What to do now

1. Inventory your AI: notetakers, CRM features, planning tools, chatbots, marketing tools.
2. Audit every AI claim on your website, ADV brochure, pitch decks and social posts. Delete anything aspirational.
3. Check vendor claims before repeating them. If a vendor says "AI-powered portfolio optimization", confirm what it does.
4. Disclose material reliance on AI where it affects advice, and document how humans review outputs.
5. Keep records of AI-generated client communications under your retention policies.

FAQ

Is there any SEC rule specific to AI for advisers now?

No AI-specific rule is in force. Existing obligations (fiduciary duty, the Marketing Rule, antifraud rules, Reg S-P and recordkeeping) apply to AI use.

Could the rule come back?

The SEC would need a new proposal and comment period. Commentators expect any future AI rule to look different from the 2023 version.

What's the fastest risk to fix?

Marketing claims. They're public, easy for examiners to check, and the basis of every AI-washing case so far.

Changelog: Oct 13, 2026: first published.