Morgan Stanley Tests GPT-4 to Speed Up Research for Financial Advisors
Morgan Stanley is piloting OpenAI's latest GPT-4 model to help its financial advisors quickly sift through massive amounts of internal research. The bank views the tool as a helpful research assistant rather than a replacement for human empathy and client interaction.
Morgan Stanley is testing an OpenAI-powered chatbot to help its financial advisors quickly navigate the bank's massive library of research and market data. Using the newly launched GPT-4 model, the tool aims to condense a process that typically takes over half an hour into a matter of seconds. Currently, 300 advisors are participating in the pilot program, which the bank plans to eventually roll out to all 16,000 advisors.
This initiative marks the first known use of ChatGPT in the banking sector. The technology acts as a digital research assistant that parses thousands of pages of intellectual capital and analyst commentary. Morgan Stanley emphasizes that the AI is designed solely to assist human workers, leaving the actual client interaction and relationship building to the advisors.
Despite the well-documented tendency of chatbots to "hallucinate" or generate convincing but incorrect information, Morgan Stanley sees real market potential in the technology. Executives acknowledge that AI lacks empathy and is just clever math, but they believe it allows advisors to spend less time digging for information and more time directly serving their clients.