What happened
JPMorgan has moved its artificial intelligence spending out of the discretionary innovation bucket and into core infrastructure, Crypto. News reported on Friday, May 9. The reclassification places the bank's $2 billion annual AI budget on the same footing as data center capacity and cybersecurity, two line items that finance teams treat as non-negotiable through the planning cycle.
Per the Crypto. News report, the change reflects how AI tooling is now wired into the bank's day-to-day operations rather than sitting in a research silo. JPMorgan has not issued a separate press release tied to the reclassification, and the disclosure surfaced through reporting rather than a formal filing.
The bank's prior commentary, including remarks from Chief Executive Jamie Dimon across 2024 and 2025 earnings calls, framed AI as a productivity lever for trading, fraud detection, and middle-office work. Treating the spend as core infrastructure formalizes that posture.
Why it matters
Reclassifying $2 billion as core infrastructure is a budgeting signal, not a product launch. It tells the market that JPMorgan's AI spend will not flex with quarterly cost discipline the way innovation budgets often do. For crypto, the read-through runs through three channels.
First, the same GPU-heavy compute stack that powers large language models is what banks lean on when they scale on-chain settlement, tokenized collateral, and real-time fraud screening for stablecoin flows. Second, JPMorgan operates Kinexys, the rebranded blockchain unit that handles tokenized deposit pilots and intraday repo, and AI-driven risk tooling sits adjacent to that infrastructure.