Treasury Unveils AI Framework Amid Calls for Enforceable Guardrails

The U.S. Department of the Treasury has launched a significant initiative to shape the responsible use of artificial intelligence within the financial sector. This move includes the phased release of multiple new resources and 'success manuals' designed to provide financial firms with practical guidance for implementing AI securely and effectively. The effort is framed as a major public-private partnership, aligning with the President’s AI Action Plan to strengthen cybersecurity and risk management. The announcement, however, is met with caution from security leaders. According to industry analysis, there are pressing calls for the Treasury's framework to include enforceable guardrails. Experts are advocating for mandatory controls covering critical areas such as adversarial testing, maintaining AI inventories, identity privilege mapping, and real-time monitoring to ensure the technology's security is robust and verifiable. This initiative represents a foundational step in establishing a common approach to AI risk. While the Treasury provides voluntary guidance and practical solutions through its collaboration with industry stakeholders, the emerging dialogue highlights a key tension between recommended best practices and the demand for legally binding security standards to protect the financial system.