The Treasury Department recently proposed recommendations for AI regulation, focusing particularly on the financial sector and broader private sector applications. Highlighting the risk of sophisticated phishing and the ability of AI to impersonate voices, the Treasury’s concerns stem from interviews with 42 finance and technology firms. These firms expressed doubts about current risk management frameworks’ ability to handle generative AI advancements, fearing a sluggish response to emerging technologies.

To address these concerns, both departments suggest implementing detailed disclosures for AI systems, akin to “nutrition labels,” to clarify the training processes and applications of these systems. They also advocate for the auditing of high-risk AI systems and establishing liability rules to determine accountability in case AI systems cause harm. While these recommendations aim to harness AI’s benefits responsibly, they are not yet mandatory. However, the Biden administration may soon issue binding guidance to further these goals, signaling a significant step towards formalizing AI oversight and fostering innovation within regulated frameworks.