JPMorgan Chase has initiated the deployment of a generative artificial intelligence (AI) tool, internally dubbed LLM Suite, that can perform tasks traditionally handled by research analysts. This new platform, likened to OpenAI’s ChatGPT, is currently available to employees within the bank’s asset and wealth management division.
According to an internal memo seen by the Financial Times, LLM Suite assists employees with writing, idea generation, and document summarization by leveraging third-party models. The memo, signed by key executives including Mary Erdoes, Teresa Heitsenrether, and Mike Urciuoli, describes LLM Suite as a virtual research analyst capable of providing information, solutions, and advice on various topics.
LLM Suite is intended for "general purpose productivity" and complements JPMorgan’s existing applications that manage sensitive financial information, such as Connect Coach and SpectrumGPT. The bank began introducing LLM Suite to select teams earlier this year, and approximately 50,000 employees, or about 15% of its workforce, now have access to the platform.
This rollout represents one of the most significant implementations of large language models (LLMs) on Wall Street. Morgan Stanley has also partnered with OpenAI to integrate AI tools into its wealth management operations.
It remains unclear if LLM Suite has experienced issues common to other AI models, such as generating inaccurate or misleading information. Nonetheless, JPMorgan is moving forward with its AI initiatives, reflecting CEO Jamie Dimon’s view that AI will transform every job, potentially eliminating some roles while creating new opportunities.
JPMorgan developed LLM Suite in-house due to regulatory restrictions preventing employees from using consumer AI chatbots like Anthropic’s Claude, OpenAI’s GPT, or Google’s Gemini. The bank's president, Daniel Pinto, has previously estimated that AI technology currently in use by the bank adds value of approximately $1 billion to $1.5 billion.
