Before presenting to JPMorgan CEO Jamie Dimon, Teresa Heitsenrether refines her message using the bank's generative AI tools. This process ensures her presentation is concise and impactful.
“I use it to extract the core message, make it clearer, and more precise,” says Heitsenrether, JPMorgan’s chief data and analytics officer, at a conference in New York. The tools, part of the bank’s broader AI strategy, are central to how she and her team streamline communication.
Dimon, a strong advocate for generative AI, eagerly awaits wider accessibility. “He’s keen to have these tools on his phone. That’s a key deliverable we’re targeting before the year’s end,” Heitsenrether shared.
Generative AI at Scale
JPMorgan, the largest U.S. bank, has deployed its generative AI assistant, the LLM Suite, to 200,000 employees, marking the beginning of AI integration across the firm. According to Heitsenrether, the current iteration focuses on improving tasks like email drafting or document summarization, but the goal is deeper integration.
“We’re looking to move from five-minute efficiency improvements to tools that save five hours,” she explains, emphasizing the transformative potential. However, achieving this vision will require time and commitment.
Adoption Sparks Healthy Competition
The LLM Suite’s launch has sparked enthusiasm and competition across JPMorgan’s divisions. Wealth and asset management led the charge by piloting an AI “copilot” for private banking earlier this year.
“When the investment bank found out, they said, ‘Wait a minute, we want that too,’” Heitsenrether recalls. This interest fuels a "flywheel effect," where demand drives innovation and adoption.
To ensure employees maximize the tools’ potential, JPMorgan offers training sessions, including how to craft effective prompts. The firm also relies on “superusers” — the top 10-20% of employees eager to champion adoption within their teams.
“We embed these superusers in different groups to act as local experts. They help their colleagues see how the tools can enhance their work,” Heitsenrether says. For instance, lawyers save hours by using AI to summarize contracts or regulations.
Overcoming Resistance
Despite growing enthusiasm, some employees remain resistant to adopting AI tools. Heitsenrether acknowledges these “pockets of resistance” but stresses the importance of addressing them early. “The sooner people engage with AI, the less skeptical they become,” she notes.
Having firsthand experience often reduces skepticism. For example, developers who use AI to draft test cases realize that the tools enhance rather than replace their work. “It demystifies AI. They see it as a partner in making their tasks more efficient,” Heitsenrether explains.
Future Vision: Personalized AI Assistants
Looking ahead, Heitsenrether envisions each employee having a personalized AI assistant tailored to their specific role. She hopes to report on progress toward this vision by next year.
Sumitra Ganesh, from JPMorgan’s AI research team, highlights ongoing pilots as foundational to these advanced systems. However, Ganesh notes that for now, human oversight is essential to ensure reliability in a heavily regulated industry.
“We don’t yet fully trust these systems,” Ganesh says. Current AI tools require human experts to verify outputs, likening the process to “training wheels” that will eventually enable greater autonomy.
A Competitive Edge for RIAs
For wealth advisors and RIAs, JPMorgan’s AI innovations signal significant opportunities. From automating client communications to streamlining research and compliance tasks, generative AI can unlock efficiencies that allow more time for personalized client interactions. As the tools evolve, advisors can expect even deeper integration, helping them stay ahead in a competitive landscape.
By embracing AI now, wealth professionals position themselves to leverage its potential fully, turning skepticism into opportunity. The shift from efficiency to empowerment is already underway, and those who adopt early will lead the industry transformation.
