The SEC’s 2027 Exam Priorities: What’s Changing for Cybersecurity and AI
On October 23, 2026, the Securities and Exchange Commission (SEC) released their exam criteria for the 2027 fiscal year. Since 2023, with the SEC’s original requirements for cybersecurity risk management, exams have focused heavily on evolving threats and putting appropriate processes in place for response; this year is particularly unique given the rise of threats related to AI*. Though the fundamentals remain the same, evolving technologies take significant precedence in the criteria. This year, the criteria highlights several key measures that firms need to be aware of, including:
- Cybersecurity, data protection, and incident response processes are mature and proactive
- Capabilities to protect data and respond to incidents are appropriately tested
- Third party risk management is a critical matter for security and business continuity
- Requirements under Regulation S-P are closely followed, especially with increasing AI processing of customer and investor data
- AI policies and technical controls are not just a “maturity demonstration” but firm requirements
Last year’s criteria involved examining the baseline processes for Reg S-P compliance. In 2027, there is both an increase of the practical controls for that compliance, as well as more oversight for controls for burgeoning technologies, like AI. Learn more about what your firm needs to do now to be ready for 2027 exams.
What Firms Should be Doing Now to Meet 2027 SEC Exam Priorities
Review existing AI policies and procedures
With AI being a major focus of controls, it is critical to ensure there are governance structures in place for such solutions. An established AI governance program is what the examiners will be looking for. In particular, it is important to have a program that describes oversight and ownership of AI practices, acceptable use of AI solutions, and risk management procedures.
The complexity of these policies and procedures will depend on the deployment of AI solutions – whether using standalone tools like Claude and ChatGPT, in-house built tools, or integrated AI solutions with existing third-party providers. The SEC has declared a particular interest in the monitoring of AI solutions for common operations (i.e. fraud prevention, trading functions, advisor recommendations) and the ability to accurately disclose the use of AI to necessary parties.
Review and test incident response plans
Last year’s enforcement of Reg S-P placed a large emphasis on development and refinement of incident response plans. By this point, such plans should be stress-tested in a controlled manner to ensure they are functional and effective. Whether this testing was conducted in a tabletop manner or through “purple team” engagements, the incident response plan needs to stand up to some kind of execution test for the SEC’s exams.
Ensuring oversight of AI tools should also be considered with incident plans. Whether that be playbooks related to investigation and response for AI solutions specifically, or involving the owners of AI processes within a firm during response activities, as AI becomes more and more engrained in firms’ operational processes, it must also be considered in security controls and response plans in turn.
Ensure third-party risks are identified, understood, and managed
Regulators across the world are increasing requirements on firms for performing thorough security risk management with respect to third-party service providers. With the exploitation of third-party vendors continuing to be a top attack vector within the financial services sector, the risks remain ever-present and significant as malicious actors increase attacks on major providers.
Recurring due diligence is a simple activity to perform to identify these potential risks as early as possible. While this is a starting step, further maturity will involve ongoing risk management procedures for vendors, and the supply chain overall, to ensure adequate protections and resiliency plans are in place.
Refine operational resiliency strategies and priorities
The SEC frames information security as a major factor in operational resiliency. In practice, meeting these requirements involves continuous feedback loops of policy and standard development, validation of control efficacy, and updates to address identified gaps.
The oft touted standards still apply of ensuring your policies and procedures for information security, business continuity, and operations are up to date, then testing the controls identified in policies via penetration testing, white box cybersecurity assessments, and gap assessments against standard frameworks such as NIST CSF.
Proof and Oversight: The Next Layer
The need to demonstrate effective controls, not just list them, remains a priority for the SEC. Policies being developed in a vacuum of technical implementations and disjointed oversight are being scrutinized more, and it is raising the burden on firms to prove that controls are not only in place, but are effective at meeting their intended purpose. The gap between indicating the presence of a data loss prevention (DLP) solution in documents, and the actual prevention of unauthorized exfiltration of data is an example of what is under the microscope in these rounds of exams.
The returning emphasis on risk management and governance practices for information security also points to the increasing interest in oversight from key figures within registered entities. Incident response planning being a key feature of recent Reg S-P updates points to the need for direct supervision during critical events related to cybersecurity, and this year’s examinations will likely reflect those updated requirements more than 2026’s audits.
The Impact of the Growth of AI on SEC Priorities
In the rush to implement the latest and greatest AI solutions, financial services firms have been, knowingly or unknowingly, increasing their risk footprint in a means to keep up. This is not an unexpected or inherently negative approach, but this rapid rise of implementation has warranted the SEC to take a deeper look at how firms are engaging with AI solutions – from the practical execution of tooling to the oversight and governance in conjunction with existing policies.
Primary features related to AI in the exams will involve assessing whether operations remain secure and in line with overlapping requirements. Bringing reference again to Reg S-P, the approaches and practices taken with processing customer and investor information with AI tools will be on the radar, as publicly available AI tools can present greater risks of exposure. AI technical controls are no longer in their infancy, so the SEC will be expecting an increase in assurance that the tools are securely implemented and checked for accuracy. Further, third-party tools increasingly leveraging AI also introduce risks, and third-party oversight with respect to AI needs to be present.
Other components of Regulation S-ID’s requirements will additionally be scrutinized with relation to AI. Given the ease in which malicious AI tools can be used to falsify identification mechanisms via voice cloning, deep fakes/live-generated video feeds, image generation, and more. The purpose of the SEC’s scrutiny is not to limit innovation and the use of emerging technologies, but to ensure the security and privacy of critical client data.
A Brief Checklist Ahead of Your 2027 SEC Audit
With the October release of the 2027 SEC criteria, firms have clear next steps for confirming their preparedness ahead of audits. To measure your readiness, confirm the following criteria are met within your firm:
- An established set of policies surrounding AI usage and any listed security controls
- Recent tests of policies, procedures, and documented plans
- A current list of critical vendors, including internal relationship owners, and latest evidence of security control validation (e.g., due diligence questionnaire)
- Recent tests of operational resiliency capabilities, like back up restoration and system failover
- Updated compliance manuals or policies covering protections of investor data and countermeasures for AI-enabled fraud attempts
How Abacus Helps Financial Services Firms
Abacus specializes in supporting financial services firms in their efforts to operationalize controls and offers regulatory-focused defensive and offensive cybersecurity services designed to help firms maintain compliance. From documented restore testing to vendor oversight workflows and AI governance guardrails, we focus on making programs easier to run and evidence when regulators ask. Contact us today to strengthen your operational readiness for 2027.
*The official statements and documents refer to “Super Intelligence” or “SI”, which is generally considered a separate selection of tools and concepts, which are not specifically segregated within the documented criteria. For the sake of simplicity, this article uses “AI” to represent the SI examination criteria.
