BNY Mellon CEO Robin Vince on Eliza, their internal AI platform powering 125+ solutions and digital employee agents
Mar 23, 2026 with Robin Vince
Key Points
- BNY Mellon has deployed 125+ AI-enabled solutions in production through Eliza, its internal multi-agentic platform built three years ago, positioning the $60 trillion custodian to externalize capabilities as services to clients.
- CEO Robin Vince frames AI as modernization across three vectors: collapsing legacy mainframe processes, enriching client products with insights, and expanding the bank's service perimeter by monetizing its data advantage.
- Vince sees tokenization of financial assets as a five-to-ten-year wave requiring infrastructure innovation beyond current blockchains, moving slower than AI adoption but offering generational value as terms and conditions embed into smart contracts.
Summary
BNY Mellon is deploying AI across three vectors: operational efficiency, product enrichment, and business expansion. CEO Robin Vince is using AI to rewrite legacy code running on mainframes and collapse multi-step processes into faster workflows. The bank is enriching client offerings with AI-powered insights and faster execution. The third vector draws inspiration from AWS. BNY Mellon custodies $60 trillion in assets globally, roughly 20% of all investable assets in the world. That scale and data throughput creates opportunities to externalize internal capabilities as services to clients, much as Amazon did with its cloud infrastructure.
Eliza is BNY Mellon's internal multi-agentic AI platform, built three years ago. The platform connects to multiple LLMs and supports both discrete AI agents and digital employees, which are multi-agent solutions wrapped with user credentials, personas, and names that operate as teammates within the organization. Vince uses Eliza daily as a CEO copilot, asking it to surface client intelligence before meetings such as recent call reports, prior interactions, relevant news, and tailored pitch ideas. The bank has deployed 125+ AI-enabled solutions in production today.
Vince sees digital assets and tokenization of stablecoins, bonds, securities, and other financial assets as the next multi-year wave, though he frames it as early-stage. He divides the arc into three generations. Gen one involves tokenizing existing assets. Gen two encodes terms and conditions into smart contracts to add real value. Gen three re-renders entire financial assets around blockchain capabilities. Current layer-one blockchains are not yet up to the task, he argues, so infrastructure innovation is required alongside adoption. This journey will take five to ten years and will move more slowly than AI adoption.
On private credit, Vince distinguishes it as distinct from public credit markets but fundamentally similar in purpose: lending into the capital system. He acknowledges recent idiosyncratic exposures and liquidity rushes but argues underlying fundamentals remain strong so long as the economy and credit markets hold. The real watch-out is energy prices. High energy costs ripple through food, manufacturing, and all energy-dependent sectors, dragging on the real economy. Employment, consumer sentiment, spending patterns, company profitability, and capital deployment are his recommended signals for tracking economic health.
Vince's broader argument is that BNY Mellon's longevity, 250 years and founded by Alexander Hamilton with the first stock traded on the NYSE, compounds advantage only if the institution adapts. AI and digital assets are not pivots but accelerants for what the bank already does at scale.