JPMorgan Chase Introduces AI Assistant Powered by OpenAI for Employees

JPMorgan Chase has introduced a new generative AI assistant, LLM Suite, to tens of thousands of its employees, marking the first phase of a broader initiative to integrate AI technology across the banking giant. The LLM Suite, powered by OpenAI’s large language models, is already supporting over 60,000 employees with tasks like writing emails, generating reports, and solving problems in Excel.

Widespread Adoption Across the Bank

The LLM Suite is expected to become as essential within JPMorgan as videoconferencing tools like Zoom, according to insiders. Rather than developing its own AI models, JPMorgan designed LLM Suite as a portal for accessing external large language models. The suite was launched using OpenAI’s model, with plans to incorporate additional models in the future.

“Ultimately, we’d like to be able to move pretty fluidly across models depending on the use cases,” said Teresa Heitsenrether, JPMorgan’s chief data and analytics officer. “The plan is not to be beholden to any one model provider.”

Generative AI’s Expanding Role

The rollout of LLM Suite at JPMorgan reflects the rapid adoption of generative AI across American corporations since the debut of ChatGPT in late 2022. JPMorgan’s move follows similar AI integrations by other industry giants, such as Morgan Stanley and Apple, signaling a broader shift towards AI-powered tools in the corporate world.

JPMorgan CEO Jamie Dimon has compared the impact of generative AI to that of electricity, the printing press, and the internet, predicting that it will “augment virtually every job” at the bank, which employs approximately 313,000 people. The bank’s cautious approach to AI, particularly in customer-facing applications, aims to mitigate risks such as the potential for chatbots to provide incorrect information.

Ensuring Data Security

Initially, JPMorgan restricted employees from using ChatGPT over concerns about exposing sensitive data to external providers. With LLM Suite, however, the bank has implemented measures to leverage the AI model while protecting its data. “Since our data is a key differentiator, we don’t want it being used to train the model. We’ve implemented it in a way that we can leverage the model while still keeping our data protected,” Heitsenrether explained.

LLM Suite is already being used across various divisions within JPMorgan, including consumer banking, investment banking, and asset and wealth management. The tool assists employees in writing, summarizing documents, generating ideas, and even creating marketing content for social media. JPMorgan’s engineers are also incorporating functions from external AI models into their programs through the suite.

The Evolution of AI at JPMorgan

JPMorgan’s journey with AI is not new; the bank has been working with traditional AI and machine learning for over a decade. However, the arrival of ChatGPT has accelerated the bank’s focus on generative AI, which offers exponentially more use cases due to its flexibility.

Heitsenrether outlined three stages in the evolution of generative AI at JPMorgan. The first involves making AI models available to employees, which is already underway. The second stage, currently being implemented, integrates proprietary JPMorgan data to enhance productivity. The third and most ambitious stage envisions generative AI powerful enough to operate as autonomous agents, significantly boosting productivity by allowing employees to focus on higher-value tasks.

Future Implications

The introduction of AI tools like LLM Suite is expected to empower workers while potentially displacing others, reshaping the banking industry in unpredictable ways. Banking jobs are particularly susceptible to automation, and AI could increase the sector’s profits by $170 billion within four years, according to Citigroup analysts.

Heitsenrether emphasized that generative AI should be seen as an assistant that handles mundane tasks, allowing employees to concentrate on more meaningful work. “You can focus on the higher-value work,” she said.