The OCC’s Super Intelligence supervisory framework (Bulletin 2026-54) and Executive Order 14434, explained
On November 4, 2026, the OCC gave its examiners a framework for reviewing how banks use what the executive branch now calls Super Intelligence. It sets no new expectations for banks, but it tells examiners to scale scrutiny to consequence, autonomy and human oversight.
What the executive order changes
Executive Order 14434, “Inaugurating the Era of Super Intelligence,” was signed on September 29, 2026 and published in the Federal Register on October 2, 2026 [3]. It is a terminology order. Section 1 states the policy that, to the maximum extent permitted by law, the executive branch shall use the terms “Super Intelligence” and “SI” in place of “Artificial Intelligence” and “AI.” Section 2 applies that to agencies’ official correspondence, public communications, websites, reports, policy documents and other non-statutory documents, and says nothing in it requires changes to previously issued regulations, Presidential actions, contracts, grants or other historical documents [3].
Section 3 supplies the definition. For purposes of the order, and except where the law provides otherwise, “Super Intelligence” and “SI” mean the technologies and systems encompassed by the term “artificial intelligence” as defined in section 9401(3) of title 15, United States Code. Section 3 also directs the Assistant to the President for Science and Technology to propose legislative language for a Federal definition within 60 days of the order [3]. Section 4 states that the order creates no enforceable right or benefit [3].
In our reading, the order’s practical effect for a bank is on vocabulary: the statutory definition it points to is unchanged, so the systems a bank already inventories as AI are the systems OCC documents now call SI.
The bulletin’s definition, word for word
Bulletin 2026-54 defines the term in its second footnote, which reads in full [1]:
- “Super Intelligence” refers to the technologies and systems previously encompassed by the term “Artificial Intelligence” as defined in 15 USC 9401(3) as well as the promise, potential, and rapidly advancing capabilities of those technologies. See Exec. Order No. 14434, 91 Fed. Reg. 63129 (Oct. 2, 2026).
The framework document repeats the same footnote [2]. The OCC’s wording adds a phrase that the order’s operative definition in section 3(a) does not contain: “the promise, potential, and rapidly advancing capabilities” comes from the order’s purpose statement in section 1 [3]. The bulletin’s first footnote defines “banks” as national banks, federal savings associations, and federal branches and agencies of foreign banking organizations [1].
What the framework tells examiners
The bulletin describes the framework as a resource to support examiners in evaluating banks’ use of SI in normal-course business activities. It states that the framework does not establish supervisory expectations for banks, that banks are not expected to use it to assess their SI use or SI risk management, and that nothing in it is intended to set enforceable standards or prescriptive requirements. Examiners are to tailor their work to the materiality of each use case, and for community banks to the bank’s size, complexity and risk profile [1].
Several points recur across the bulletin and the framework [1] [2]:
- SI is not a separate risk category. It can introduce, amplify or modify risk in existing categories such as operational, compliance and credit risk, and those risks are to be handled through the bank’s existing risk management framework.
- Control intensity could increase with the potential consequences of an SI agent’s failure or misuse.
- Risk classification may rest on potential consequences, transfer of agency and the degree of human oversight, rather than on the technical sophistication of the SI. The framework’s example: a simple rules-based agent authorized to initiate transactions could present greater risk than a sophisticated model limited to internal drafting assistance.
- Model risk management principles described in OCC Bulletin 2026-13 apply to non-generative and non-agentic SI models, so the framework is more applicable to generative and agentic SI.
Six pillars and sixteen components
The framework groups its sixteen components under six pillars: purpose (the actions a system or agent is permitted to execute), risk (the consequences if an output or action is not fit for purpose), control, oversight, resilience and assurance [2]. The components run from defined purpose and permitted use, risk classification and human oversight through governance, model risk management, data governance, privacy, compliance, third-party risk, operational resilience, cybersecurity, performance monitoring, testing, documentation, internal audit, and reporting [2].
An illustrative four-tier scale
Section 3 of the framework describes an illustrative method that may score each use case on decision consequence, reversibility, customer and financial impact, autonomy, data sensitivity, explainability and provenance, and combine the scores into a composite tier [2]:
- Tier 1, critical: direct action, low or no reversibility, high impact, autonomous or partially autonomous. Examples: an autonomous payment agent, real-time credit decisioning, fraud-hold release, market access.
- Tier 2, high: direct action or high-weight decision support with a human on the loop. Examples: SI-assisted underwriting, agent-initiated customer communications, AML alert disposition.
- Tier 3, moderate: decision support with a mandatory human in the loop, limited data sensitivity, reversible. Examples: summarizing documents for credit memos, agent-assisted research.
- Tier 4, low: no customer or financial data exposure, no direct action, fully human-controlled. Examples: internal productivity tools, general employee Q&A.
The framework says a fully autonomous, customer-facing payment agent is Tier 1 regardless of how well it performs in testing [2].
Where examiners are likely to look
The framework’s language is largely permissive (“may,” “could”), so the following are its illustrative considerations, not requirements [2]:
- A written purpose for each system or agent: what it is and is not designed to do, who may invoke it, the data, systems and actions it may use, and when human review or suspension applies. Activity outside the approved purpose may be treated as a control failure even if the result looks reasonable.
- Human oversight that can be shown to work: the framework distinguishes human-in-the-loop (review and approval before an output is used or an action executed) from human-on-the-loop (active monitoring with authority and practical ability to intervene), and warns that high volumes or bulk approval could reduce review quality and contribute to alert fatigue.
- Testing that fits non-deterministic systems: repeated testing, defined samples and consistency thresholds may give better evidence than a single successful run.
- Vendors: development or hosting may be spread across vendors, but accountability for the bank’s risk generally remains with the bank.
One tension is worth reading closely. The bulletin says banks are not required to define a separate SI risk appetite, while the governance component says the board and senior management may establish an SI risk appetite [1] [2]. Read together, an appetite statement that covers SI is an option the framework describes, not an obligation.
Status as of November 14, 2026
The OCC says it intends to update the guide as technology, use cases and supervisory understanding mature [1]. The framework’s third footnote states that it was prepared with the assistance of SI and that the content was reviewed, edited and approved by a human [2]. The OCC’s model risk guidance that the framework points to is covered in our explainer on the 2026 interagency model risk management guidance.
This explainer describes the bulletin, the framework and the order; it is not legal advice, and banks should confirm with counsel how they apply.
Moraph’s teams help organizations with work like this. See our solutions.
Spotted an error? Tell us and we will check it and correct the explainer where needed.