SEC AI Examination Priorities for 2026
The SEC's Division of Examinations has made AI governance a priority for 2026. If you're a registered investment adviser, broker-dealer, or fund, expect questions about how your firm uses and governs AI tools.
What Examiners Will Ask
Based on SEC guidance and recent enforcement actions, expect these areas of inquiry:
AI Tool Inventory
- What AI tools does your firm use?
- Who approved them?
- What data do they access?
- Where is the documentation?
Policies and Procedures
- Do you have an AI acceptable use policy?
- How are new AI tools evaluated before adoption?
- Who is responsible for AI governance?
- How do you handle unapproved tools?
Client Data Protection
- Which AI tools access client PII or financial data?
- What are the vendor's data handling practices?
- Is client data used for model training?
- Where is data processed and stored?
Conflicts of Interest
- Do AI tools create conflicts (e.g., AI recommending products where the firm has a financial interest)?
- How are AI-generated recommendations reviewed before reaching clients?
- Is there human oversight of AI-driven decisions?
Disclosures
- Have you disclosed AI use to clients where required?
- Are marketing materials generated by AI reviewed for compliance?
- Do clients know when they're interacting with AI?
How to Prepare
Step 1: Inventory
Create a complete registry of every AI tool used across the firm. Include: tool name, vendor, use case, data access, approval status, responsible person.
Step 2: Policies
Implement an AI acceptable use policy covering: approved tools, prohibited uses, data classification, approval process, monitoring.
Step 3: Process
Build an auditable approval workflow for new AI tool requests. Every request, review, and decision should be logged with timestamps and rationale.
Step 4: Audit Trail
Ensure you can produce documentation on demand. Examiners won't accept "we handle it informally."
Step 5: Ongoing Monitoring
AI governance isn't set-and-forget. Monitor approved tools for vendor changes, new risks, and compliance drift.
ShieldAI automates SEC examination readiness for AI governance →