JPMorgan’s AI stack centers on a proprietary LLM Suite, used by more than 230,000 employees to integrate internal data and external information via specialized agents, with over 500 active AI use cases in production across fraud detection, investment banking deck generation, compliance review, and predictive liquidity management for corporate treasurers, all operating on scalable Microsoft Azure and Snowflake infrastructure.

Fraud-related AI has driven a 95% reduction in false positives for anti-money laundering, illustrating the tangible impact of the bank’s AI deployment across its operations.

The move signals to the banking industry that AI is a non-discretionary cost at scale, potentially pushing competitors to treat AI as essential infrastructure rather than a discretionary project.

JPMorgan reclassified its AI investment from discretionary innovation to core infrastructure, committing $2 billion annually and positioning it alongside data centers, payment systems, and cybersecurity within a $19.8 billion technology budget for 2026.

Dimon frames the AI and blockchain play as JPMorgan’s primary competitive moat amid rising stablecoin threats and economic uncertainty, with OpenAI and other AI-native players intensifying competition targeting institutional clients.

The reclassification signals to the market that AI is a non-discretionary cost, and modernization spending is shifting toward products, platforms, and AI integration as baseline operating costs.

JPMorgan is advancing digital assets with JPMD, integrating AI to manage flows and predict institutional liquidity on public blockchain infrastructure, signaling a convergence of AI infrastructure with digital asset rails.

CEO Jamie Dimon says the AI program has self-funded through about $2 billion in operational savings across more than 150,000 employees, delivering roughly 10% to 11% productivity gains in engineering, operations, and fraud detection.

AI-powered real-time monitoring of transactions on scalable, governed infrastructure underpins fraud detection improvements, contributing to a strong reduction in false positives and enhanced AML vigilance.