The SEC’s 2026 examination priorities place AI governance at the center of the exam agenda for registered investment advisers and broker-dealers. If your firm uses algorithmic tools for portfolio construction, trade execution, client communications, or surveillance, examiners are coming with specific questions about how you supervise those tools. This post explains what the SEC 2026 examination priorities AI focus means in practice, where IA and broker-dealer obligations diverge, and what CCOs and CROs should do before the exam cycle begins.
Why AI Tops the SEC’s 2026 Examination Priorities List
The SEC’s Division of Examinations has signaled for two consecutive years that artificial intelligence represents a systemic risk to investor protection — not a theoretical one. The 2025 examination priorities document named AI as an emerging area of focus; the 2026 cycle elevates it to a primary risk theme across the IA and broker-dealer programs. AI adoption in financial services accelerated faster than compliance infrastructure kept pace. Firms deployed large language models for client-facing communications, machine learning models for trade recommendations, and automated surveillance tools — often without updating their written supervisory procedures, conducting formal model validation, or documenting how those tools interact with fiduciary obligations. The SEC’s examination priorities list for 2026 reflects that gap. Three factors explain why AI sits among the leading focus areas on the SEC AI risk examination focus for 2026:
- Breadth of adoption. Unlike digital assets, which remain concentrated in a subset of firms, AI tools are now embedded across much of the industry — from large wirehouses to smaller RIAs using off-the-shelf portfolio analytics platforms. The examination program follows the distribution of risk.
- Investor harm potential. AI-generated investment recommendations that are biased, unexplainable, or untested create direct investor harm pathways. The SEC has cited the potential for algorithmic tools to generate unsuitable recommendations at scale as a core concern.
- Regulatory gap. The AI governance regulatory landscape 2024 had no single statute equivalent to the Investment Advisers Act’s fiduciary standard or FINRA’s suitability rules, and that gap persists into 2026. Examiners are applying existing frameworks — the Advisers Act, Regulation Best Interest, the Marketing Rule — to AI use cases, and the 2026 priorities signal that firms should not wait for new rules before building controls.
What the SEC’s 2026 AI Examination Priorities Mean for Investment Advisers
For RIA CCOs, the SEC IA exam program AI priorities translate into concrete compliance obligations that map to existing regulatory frameworks rather than new ones.
- Fiduciary duty and AI-assisted recommendations. When an investment adviser uses an AI tool to generate, filter, or rank investment recommendations, the fiduciary obligation does not transfer to the algorithm. Examiners will assess whether the firm has documented how AI outputs are reviewed before they influence client advice, and whether advisers understand the basis for AI-generated recommendations well enough to exercise independent judgment.
- The Marketing Rule and AI-generated content. SEC guidance on AI for investment advisers has consistently flagged the Marketing Rule as a high-risk intersection. AI-generated performance presentations, testimonials, and client communications must meet the same substantiation and disclosure standards as human-authored content. Firms using generative AI for marketing materials without a documented review and approval workflow are exposed.
- Compliance program adequacy. Rule 206(4)-7 requires RIAs to maintain written policies and procedures reasonably designed to prevent violations. Examiners applying SEC AI governance guidance for 2025 and 2026 will assess whether those policies specifically address AI tool selection, ongoing monitoring, and incident response. A generic technology use policy is unlikely to satisfy an examiner who arrives with an AI-specific request list.
- Conflicts of interest. If an AI tool is provided by a third party that also has a financial relationship with the adviser — a common arrangement with portfolio analytics and model marketplace vendors — examiners will look for conflict disclosure and management procedures specific to that arrangement.
CCOs at mid-market RIAs should treat AI governance as a compliance program amendment project, not a technology project. The AI Governance Framework for Investment Advisers: Meeting SEC Examination Standards provides a structured starting point for that work.
Broker-Dealer AI Compliance: How the 2026 Priorities Differ from IA Requirements
Dual-registrants and standalone broker-dealers face a distinct set of SEC examination priorities for broker-dealer AI that diverge from the IA program in three important ways.
- Regulation Best Interest and algorithmic suitability. Reg BI requires broker-dealers to act in the best interest of retail customers at the time of a recommendation. When a recommendation is generated or materially influenced by an AI model, the broker-dealer must demonstrate that the model’s outputs are consistent with Reg BI’s care obligation — including that the model accounts for the customer’s investment profile and does not systematically favor higher-cost products. This is a different analytical frame than the RIA fiduciary standard, and examiners will apply it differently.
- FINRA Rule 3110 and supervisory systems. Broker-dealers are subject to FINRA’s supervisory rules, which require written supervisory procedures and supervisory control systems. SEC broker-dealer AI compliance guidance, read alongside FINRA’s own AI guidance, requires that AI tools used in the supervisory process — for example, automated surveillance of communications or trade monitoring — are themselves subject to supervision. Firms cannot use an AI tool to satisfy a supervisory obligation without also supervising the tool.
- Books and records. Broker-dealers face more prescriptive recordkeeping obligations than RIAs under Exchange Act Rules 17a-3 and 17a-4. When AI tools generate, modify, or route records — including communications, order records, or surveillance alerts — firms must ensure those records are captured in compliant formats and retention schedules. The SEC exam focus areas for 2026 fintech include specific attention to whether AI-generated records meet Exchange Act standards.
- Third-party vendor risk. Both programs address vendor risk, but broker-dealers operating under FINRA oversight face additional expectations around vendor due diligence documentation. Examiners will ask for evidence that the firm assessed the AI vendor’s model validation practices, data governance, and incident history before deployment — not just a vendor contract with standard representations.
For firms registered as both an RIA and a broker-dealer, the 2026 examination cycle creates a compounding exposure: examiners from both the IA and BD programs may review the same AI tools under different regulatory lenses. Understanding the Multi-Regulator AI Compliance for Financial Institutions: SEC, OCC, NYDFS, CFPB, and Federal Reserve framework helps dual-registrants map overlapping obligations without duplicating effort.
How SEC Examiners Will Assess AI Governance: Documentation, Controls, and Audit Trails
The SEC’s Division of Examinations focus on AI has historically been operationalized through document requests and on-site interviews. The 2026 AI examination program follows that pattern, with a specific set of artifacts examiners are likely to request.
- Inventory of AI tools. Examiners will ask for a complete list of AI and algorithmic tools in use, including vendor-provided tools embedded in third-party platforms. Firms that cannot produce a current inventory signal immediately that governance is informal.
- Written policies and procedures. The SEC examination manual AI section, as applied by staff in recent examinations, looks for policies that address: how AI tools are selected and approved, how they are monitored after deployment, how errors or unexpected outputs are escalated, and who is accountable for ongoing monitoring.
- Model validation and testing records. For quantitative models used in investment recommendations or surveillance, examiners will look for evidence of pre-deployment testing, ongoing performance monitoring, and periodic revalidation. This is an area where many mid-market firms have significant gaps — the tools are running, but the validation documentation does not exist.
- Audit trails. When an AI tool influences a client recommendation, a trade, or a compliance determination, examiners want to see a record that connects the AI output to the human decision. The SEC exam priorities for 2026 artificial intelligence specifically contemplate scenarios where firms cannot reconstruct how a recommendation was generated — a finding that supports a deficiency letter on its own.
- Training records. Examiners will ask whether supervised persons who use AI tools have received training on their limitations, appropriate use cases, and escalation procedures. A gap here is easy to remediate before an exam; it is harder to explain after.
The What SEC Examiners Will Ask About Your AI Controls: Documentation and Audit Trail Requirements post goes deeper on each of these artifact categories with specific examples of what adequate documentation looks like.
How to Position Your Firm Before the 2026 Exam Cycle Begins
The SEC’s examination calendar means most mid-market RIAs and broker-dealers will face AI-focused examinations in 2026 or early 2027. Firms that begin preparation now have a meaningful advantage — not because they can game the process, but because the governance work takes time to complete and document credibly.
- Complete an AI tool inventory first. Before you can govern AI, you need to know what you have. This includes tools your technology team deployed, tools embedded in vendor platforms, and tools individual advisers or traders adopted informally. The inventory is the foundation for every other governance step.
- Map tools to regulatory obligations. For each tool, identify which regulatory frameworks apply — fiduciary duty, Reg BI, the Marketing Rule, FINRA supervisory rules, Exchange Act recordkeeping — and document the analysis. This mapping exercise surfaces gaps and prioritizes remediation work.
- Update written supervisory procedures. Generic technology policies will not satisfy examiners. Procedures should address AI-specific scenarios: how AI-generated recommendations are reviewed, how AI-assisted communications are approved, how model errors are escalated, and who is accountable for ongoing monitoring. The CCO and CRO Guide to AI Governance: Roles, Responsibilities, and SEC Exam Readiness addresses how to structure internal accountability so that AI governance does not fall into the gap between compliance and technology.
- Run a pre-examination readiness assessment. Before examiners arrive, test your documentation against the request list they are likely to bring. The How to Prepare for a SEC AI Governance Examination: Checklist for Investment Advisers and Broker-Dealers provides a structured checklist for that exercise.
The firms that will fare best in the 2026 AI examination cycle are not necessarily the ones with the most sophisticated AI tools — they are the ones that can demonstrate they understand what their tools do, have documented controls around them, and have assigned clear accountability for ongoing oversight. That is achievable for mid-market firms that start now.
- Ready to assess your firm’s SEC 2026 AI examination readiness? Brine offers a structured readiness assessment and platform demonstration scoped specifically to the 2026 AI examination priorities for mid-market RIAs and broker-dealers. See where your governance program stands before examiners do.
- This post is part of the pillar: SEC AI Examination Priorities for Investment Advisers
- Related reading: How to Prepare for a SEC AI Governance Examination: Checklist for Investment Advisers and Broker-Dealers · AI Governance Framework for Investment Advisers: Meeting SEC Examination Standards · What SEC Examiners Will Ask About Your AI Controls: Documentation and Audit Trail Requirements · CCO and CRO Guide to AI Governance: Roles, Responsibilities, and SEC Exam Readiness*