The Securities and Exchange Commission has placed artificial intelligence inside its fiscal 2027 examination agenda, making the technology part of the supervisory conversation for investment advisers, broker-dealers and other regulated market participants. The published priorities also emphasize cybersecurity, operational resilience and the risks created by increasingly complex technology stacks.
Examiners will look past the AI label
The agenda signals attention to whether firms’ controls match how they actually use emerging technology. That can include governance around automated recommendations, protection of customer information, supervision of third-party services and the accuracy of representations made to clients about AI-enabled products.
Operational resilience matters because a model or data service can fail even when its outputs are not fraudulent. Regulators can examine business-continuity planning, incident response, access controls and whether dependencies on vendors create risks that a firm has not identified or tested.
A supervisory map, not a new rule
Examination priorities tell firms where SEC staff expects to focus during reviews. They do not amend securities law, create a separate AI licensing regime or establish that any named practice is unlawful. Findings depend on each firm’s facts and its existing duties.
For compliance teams, the practical message is to connect AI projects to established obligations rather than treating them as an isolated innovation program. Documentation, testing, human oversight, vendor diligence and truthful disclosures remain the core evidence examiners are likely to expect.