What's Happening?
The Office of the Comptroller of the Currency (OCC) has released a new supervisory framework to guide its examiners in evaluating national banks' use of Super Intelligence (SI) in their normal business operations. This framework is designed to assist
examiners in assessing the risks and controls associated with banks deploying SI and agentic SI, which are rapidly evolving technologies. The OCC emphasizes that this framework does not establish new supervisory expectations for banks but serves as a resource for examiners to tailor their evaluations based on the materiality, nature, and potential consequences of a bank's SI use cases. The guidance clarifies that examiners should not treat SI as a separate risk category, nor are banks required to define a distinct SI risk appetite. Instead, the framework suggests that a bank's control intensity should increase with the potential consequences of an SI agent's failure or misuse, and risk classification may be based on potential consequences, transfer of agency, and human oversight rather than technical sophistication.
Why It's Important?
This supervisory framework marks a significant step for the U.S. financial industry as it navigates the integration of advanced artificial intelligence technologies. By providing guidance to examiners, the OCC aims to foster responsible innovation while ensuring the safety and soundness of banking operations. The framework's emphasis on tailoring examinations to a bank's size, complexity, and risk profile, proportionate to its SI use, is crucial for community banks, preventing them from being subjected to overly burdensome regulations designed for larger institutions. The OCC's technology-neutral approach signals its intent to encourage innovation without imposing rigid, prescriptive requirements that could stifle technological advancement. This guidance helps banks understand how their SI implementations will be assessed, promoting a proactive approach to risk management and governance in an area of rapid technological change. It also highlights the OCC's recognition that SI, while transformative, introduces unique considerations that need to be integrated into existing risk management practices.
What's Next?
The OCC intends to regularly update this supervisory framework to adapt to the continuous evolution of SI technology, new use cases, and feedback from stakeholders. This iterative approach suggests that the regulatory landscape for AI in banking will remain dynamic, requiring banks to stay agile in their compliance and risk management strategies. As banks continue to deploy SI across various functions, examiners will use this framework to identify SI-specific considerations that complement existing model risk management, governance, technology, compliance, and operational-risk practices. The ongoing development of this framework will be critical in shaping how financial institutions in the U.S. adopt and manage AI, balancing the benefits of innovation with the imperative of maintaining financial stability and consumer protection. Banks will need to closely monitor these updates and adjust their internal policies and procedures accordingly.
Beyond the Headlines
The OCC's framework for Super Intelligence reflects a broader trend among financial regulators globally to address the implications of AI. While not establishing new regulatory expectations, the guidance implicitly encourages banks to integrate AI risk management into their existing frameworks, rather than treating it as an isolated concern. This approach acknowledges the pervasive nature of AI and its potential to impact various aspects of banking, from operational efficiencies to customer service and fraud detection. The focus on 'control intensity' based on potential consequences, rather than technical complexity, suggests a pragmatic regulatory philosophy that prioritizes outcomes and impacts. This could lead to a more robust and adaptable risk management culture within banks, where the ethical, legal, and operational implications of AI are considered holistically. The framework also subtly encourages a shift towards greater human oversight and accountability in AI-driven processes, recognizing the limitations and potential biases inherent in advanced AI systems.
