Leaders of the United States Federal Reserve are expressing confidence in the potential of generative artificial intelligence (AI) tools to act as a “super analyst” for banks and the government. These tools have the capability to handle customer service tasks for banks and even replace human programmers.
Sunayna Tuteja, the Federal Reserve’s chief innovation officer, recently participated in a fireside chat at the Chicago AI Week event with Margaret Riley, a senior vice president at the Fed’s financial services division. The discussion centered around the advancement of responsible AI innovation within the Federal Reserve System.
According to a report from Risk.net, Tuteja and Riley explored five potential use cases for generative AI being explored by the Fed. These include data cleansing, customer engagement, content generation, translating legacy code, and improving operational efficiency.
Riley described the overall potential of generative AI as that of a “super analyst” that can simplify tasks for Fed employees and serve as a customer support specialist, enhancing banks’ ability to interact with clients on a personalized level.
When discussing the translation of legacy code, Tuteja expressed a belief that large language models (LLMs) like ChatGPT or similar AI products could potentially replace certain roles traditionally carried out by humans.
However, both Tuteja and Riley emphasized that generative AI and LLMs have their limitations. The discussed use cases are still in the exploratory phase and caution must be exercised. While the risks of implementing generative AI systems in sectors where accuracy is crucial, such as finance, are well-documented, Tuteja issued a strong warning about the potential drawbacks of not embracing these technologies.
In related news, it is estimated that losses resulting from cryptocurrency and deep fake scams could reach $25 billion by 2024, according to Bitget.