The news: JPMorgan is testing agentic AI systems that make portfolio allocation decisions, marking a shift in AI use from an assistant to a decision-maker, per Bloomberg. In historical backtests, the bank's best-performing AI agent outperformed a traditional 60/40 stock-bond portfolio and its own rules-based allocation model while delivering lower volatility. 

Zooming in: JPMorgan built a series of AI agents using models from OpenAI and Anthropic that classify markets into different economic regimes—such as Goldilocks, reflation, stagflation, and risk-off—and determine how to allocate assets based on those conditions. All AI agents it tested outperformed the bank's existing rules-based framework on a risk-adjusted basis. 

At the same time, JPMorgan emphasized that the technology should complement—not replace—human expertise, warning that historical backtests are not evidence the models will consistently outperform in live markets. It also acknowledged risks such as crowded trading strategies, market manipulation, and overconfidence in AI-generated recommendations.

Why it matters: JPMorgan's experiment builds on a broader emerging trend across wealth management. Earlier this year, Charles Schwab expanded its AI-generated financial guidance to everyday investors, using genAI to make personalized advice more accessible. JPMorgan's work extends beyond customer-facing recommendations to institutional portfolio construction, as AI is beginning to influence how financial advice is delivered and investment decisions are made. 

As AI capabilities mature, competitive differentiation may hinge on firms' ability to combine proprietary data, investment expertise, and strong governance with more autonomous AI systems.

Implications for banks: The initiative signals that large financial institutions are beginning to experiment with AI for higher-value, judgment-based investment decisions rather than simply automating more manual tasks.

JPMorgan's experiment foreshadows the next phase of AI adoption across financial services and raises the competitive bar for banks and asset managers. Indeed, firms with more sophisticated AI capabilities could potentially improve investment performance, personalize portfolio management at scale, and reduce reliance on static rules-based models. This also increases pressure on competitors to invest in agentic AI while developing governance, risk management, and human oversight to deploy systems responsibly. 

This content is part of EMARKETER’s subscription Briefings, where we pair daily updates with data and analysis from forecasts and research reports. Our Briefings prepare you to start your day informed, to provide critical insights in an important meeting, and to understand the context of what’s happening in your industry. Non-clients can click here to get a demo of our full platform and coverage.