JPMorgan Chase unveils AI-powered LLM Suite; may replace research analysts

US financial services major JPMorgan Chase's new large language model, functioning as a 'research analyst', has already been rolled for more than 50,000 employees

JPMorgan Chase has unveiled a generative artificial intelligence product, dubbed LLM Suite, designed to enhance the productivity of its employees in the asset and wealth management division, according to a report by the Financial Times (FT).

The large language model (LLM) tool is reminiscent of OpenAI’s ChatGPT, which aims to function as a ‘virtual research analyst’, offering information, solutions, and advice on various topics.

Key features of the AI-powered LLM suite include:
 

Writing assistance: Helps employees generate and refine written content.

Idea generation: Assists in brainstorming and developing new concepts.

Document summarisation: Provides concise summaries of lengthy documents.

The LLM Suite could play a crucial role in improving general productivity, and complements JPMorgan’s other apps Connect Coach and SpectrumGPT, which manage sensitive financial information.

 

JPMorgan block ChatGPT, develops own AI system

Last year, the US financial services major had restricted employees from using OpenAI’s ChatGPT at the peak of the AI platform’s buzz. At the time, the bank did not specify any reason for the restriction or how the artificial intelligence tool may have been used by employees. Reports had also indicated that JPMorgan and other organisations had gone as far as to block access to ChatGPT.

By May 2023, however, CNBC reported that JPMorgan Chase was developing its own AI software, similar to ChatGPT. The speculation arose after the bank applied for a trademark for a product called IndexGPT.

“It’s an AI programme to select financial securities,” trademark lawyer Josh Gerben told the news company. “This sounds to me like they’re trying to put my financial advisor out of business,” Gerben added.

The CNBC report also noted that trademarks can take up to a year to be approved, following which the bank has three years to launch the product, to secure the trademark.

According to FT, the bank’s decision to develop LLM Suite in-house stems from strict regulatory requirements preventing the use of consumer AI chatbots, such as OpenAI’s GPT or Google’s Gemini, due to concerns over client data security.

Wall Street adopts AI

Since the initial introduction of LLM Suite earlier this year, approximately 50,000 employees, representing about 15 per cent of JPMorgan’s workforce, have gained access to the software. This rollout marks one of the most significant in-house applications of a large language model in the financial sector.
 

The move also aligns with the broader trend on Wall Street, with Morgan Stanley also leveraging AI through a partnership with OpenAI.

CEO Jamie Dimon has highlighted AI’s transformative potential, suggesting it will revolutionise every job within the bank, potentially eliminating some roles while creating new opportunities. President Daniel Pinto estimates that the value derived from AI technologies currently used by JPMorgan ranges from $1 billion to $1.5 billion.