Leaders at the United States Federal Reserve view generative artificial intelligence (AI) as a “super analyst” that can significantly enhance the agency’s efficiency and capabilities. The Fed’s chief innovation officer, Sunayna Tuteja, discussed this perspective during the Chicago AI Week event in a conversation with Margaret Riley, SVP at the Fed’s financial services division. They focused on “Advancing responsible AI Innovation at the Federal Reserve System.”
Tuteja and Riley identified five key use cases for generative AI at the Fed: data cleansing, customer engagement, content generation, translating legacy code, and enhancing operational efficiency. Riley highlighted the potential of AI to act as a “super analyst,” easing the workload for Fed employees and improving customer support for banks by offering more personalized and efficient service.
One significant area of discussion was the role of AI in translating legacy code. Tuteja suggested that large language models (LLMs), such as ChatGPT, could replace some traditional coding jobs. She noted that rather than hiring developers to update old code, LLMs could be leveraged for this task, allowing developers to act more as auditors or editors.
Despite the enthusiasm for AI, Tuteja and Riley acknowledged its limitations and emphasized that these use cases are still exploratory. Tuteja warned that while it’s important to consider the risks of implementing new AI systems, it’s equally crucial to recognize the potential risks of inaction. In some cases, not adopting new technologies could pose greater risks than trying to integrate them responsibly.