AI Strategy Index

BNY AI Strategy: AI for Everyone, With Digital Employees That Have Managers

How BNY approaches AI: Eliza for every employee, 99% trained, more than 140 digital employees with human managers and over 400 AI solutions in production. Updated November 2026.

Company
BNY
Industry
Banking & financial services
Headquarters
New York, United States
Region
US
AI strategy
Published AI strategy
Index group
S&P 100

Custody bank and asset servicer: securities services, market and wealth services, and investment and wealth management

Where the AI strategy is set out

BNY paper “AI in the next phase: driving adoption” (October 2025) sets out its approach: AI for everyone, everywhere and everything. It sets no targets.

Read the company source

Where BNY Is Going with AI

BNY wants AI used by everyone, everywhere, for everything, and treats it as a way to create capacity. Nearly every employee has been trained on Eliza, its in-house AI platform, about half use AI every day, and more than 140 “digital employees” work alongside people, each with a human manager. BNY had more than 400 AI solutions in production by October 2026 and reports speed gains use case by use case. Headcount fell to about 46,600 in September 2026.

AI is a capacity generator. Chairman and CEO Robin Vince said in October 2026: “At BNY, AI is for everyone, everywhere and everything” and “We think about it as a capacity generator” (third-quarter 2026 call). He described three steps: create capacity inside BNY, improve its products, then build new offerings for clients. The first is where BNY started, and the work is now shifting to the second. In July 2026 he said: “adoption and embedding in a company is going to be the differentiator” (American Banker).

The focus has four parts.

  • Every employee: Eliza, an AI platform that connects models from OpenAI, Google and Anthropic to BNY’s data, and on which more than half of employees build their own agents (first-quarter 2026 call, via PYMNTS)
  • Digital employees: more than 140 autonomous AI workers (Axios, August 2026), some working around the clock in payment operations (InvestmentNews)
  • Engineering: about 40% of code is written by AI (April 2026)
  • Clients: over time, BNY plans to extend its AI capabilities to clients (October 2026)

What It Takes: One Platform, Training for Everyone and Agents With Managers

One platform for the whole bank. Eliza, launched in 2024, gives employees a single, controlled place to use AI and build agents (Fortune). In December 2025, BNY connected Google’s Gemini Enterprise to it. Chief Data and AI Officer Sarthak Pattanaik said: “At BNY, our AI strategy is clear: AI for everyone, everywhere, and everything” (CDO Magazine). Vince says Eliza routes each task to a suitable model and manages the cost, which only works at scale (October 2026). In September 2026, BNY opened an AI lab at Carnegie Mellon University to research how to check AI-written code and test agents for critical systems (October 2026).

Training for everyone. In September 2025, the chief people officer and the chief information officer told staff: “Fluency in AI is no longer optional, it’s critical” (Fortune). By October 2025, 99% of employees had been trained and onboarded, through role-based learning paths, a responsible AI curriculum, bootcamps and AI champions (BNY, October 2025). Nearly all employees have completed ten hours of Eliza training (InvestmentNews).

Agents get managers. Each digital employee has a persona, its own access credentials and a human supervisor, and AI solutions pass risk, data and security reviews before release (BNY). Vince said: “We don’t fully trust even people to do things solely by themselves”, so agents work under the same maker-checker controls as people (October 2026). It is a working example of hybrid teams of people and AI agents, and of why agents need managers.

Headcount is falling. BNY had about 53,400 employees at the end of 2023 and about 46,600 in September 2026. Vince argues that AI savings pay for growth: “I think when you can save in one place, it allows you to be able to do more” (American Banker, May 2026). He calls AI a “capacity creator” and told incoming analysts to “experiment with technology like AI” (Axios), a live case of AI and entry-level jobs. BNY has not said how much of the decline is due to AI.

AI fluency, coded against the Lead with AI Fluency Matrix (September 2026 edition):

  • Fluency floor: Level 1 (Chatting), backed by near-universal training. 99% of employees have been trained and about half use AI every day.
  • Highest level evidenced: Level 3 (Integrating), company-claimed. AI is built into onboarding, compliance, coding and account planning, with measured speed gains. Digital employees work in areas such as payment operations under human managers, but BNY has published no result for a process they run end to end, so Level 4 is not evidenced.
  • Portfolio design: Platform-led, with every employee a builder. One platform serves the whole bank, digital employees have human managers, and the work moves from internal capacity to products to client offerings.

How BNY Measures Results

BNY reports results by use case. All figures below are company-claimed:

  • Scale: more than 400 AI solutions in production in October 2026 (third-quarter call), up from 117 in September 2025 (American Banker)
  • Speed: client onboarding more than 20% faster, compliance cases resolved more than 30% faster, and account plans drafted with AI completed 60% faster (April 2026)
  • Engineering: about 40% of code written by AI, contributing to a double-digit rise in how fast software is released (April 2026)
  • Adoption: about half of employees use AI daily, and more than half build agents (April 2026)

There is no total. BNY spends about $4 billion a year on technology but does not break out AI spending, which Vince said could be counterproductive. CFO Dermot McDonogh said BNY has “harvested efficiency gains that we’ve reinvested in the business” (October 2026), without a figure for AI. Putting a number on it is the step from AI adoption to AI value.

Next markers: a figure for AI savings, the first AI offerings for clients, results from digital employees, and whether revenue keeps growing as headcount falls.

Change log

  • 2024: BNY launches Eliza.
  • September 2025: A new version of Eliza lets employees build agents; staff are told AI fluency is no longer optional.
  • October 2025: 99% of employees are trained, and 117 AI solutions are in production.
  • December 2025: Google’s Gemini Enterprise is connected to Eliza.
  • April 2026: About 220 AI solutions and about 140 digital employees are in production.
  • September 2026: The BNY AI lab at Carnegie Mellon University opens.
  • October 2026: More than 400 AI solutions are in production.

The Bottom Line

  • Disclosure: 4 of 4. BNY answers all four questions, with results reported use case by use case.
  • AI is everyone’s job, with near-universal training and about half of employees using AI daily.
  • Agents are managed like staff, with credentials, a supervisor and the same controls as people.
  • Results come as speed gains, with no total and no separate AI budget.
  • The say-do gap is headcount. Vince calls AI a capacity creator, yet headcount fell about 13% from the end of 2023 to September 2026, and BNY does not say how much of that is AI.
  • Watch the move to clients. The test is whether AI turns from internal capacity into products clients pay for.

Revenue, Headcount and Revenue per Employee

What the numbers show: Revenue rose 21% from 2022 to 2025 while headcount fell 7%, so revenue per employee rose 31%, to about $417k. Headcount kept falling in 2026, to about 46,600 in September, and BNY says AI is helping it raise revenue per employee. These figures do not show that AI caused any change.

Revenue, headcount and revenue per employee, FY2022 to FY2025

  • FY2022: $16.53bn revenue, 51,700 employees, $320k per employee
  • FY2023: $17.70bn revenue, 53,400 employees, $331k per employee
  • FY2024: $18.62bn revenue, 51,800 employees, $359k per employee
  • FY2025: $20.08bn revenue, 48,100 employees, $417k per employee

Change, FY2022 to FY2025: revenue up 21%, headcount down 7%, revenue per employee up 31%.

Notes: Revenue is total revenue. 2022 and 2023 are as restated in 2024 for a new accounting standard on tax credit investments (ASU 2023-02), which raised them by $152 million and $195 million against first reports. Headcount is approximate, at year-end. Headcount kept falling in 2026, to about 46,600 at the end of September.

These figures give context on scale and productivity. They do not show that AI caused any change. Sources: BNY 10-K filings and 2024 revised financial supplement.

How BNY Compares with Financial Services Peers

We looked at how BNY compares with other financial services companies in the index. Here is how they stack up.

CompanyDisclosureAI ownerFluency floorGovernanceAI value reportedRevenue per employee
Allianz3 of 4No named AI owner; board COO covers ITLevel 1, by participationEight principles and a register of every use caseResults by use case, no total$1.35M
American Express3 of 4EVP, Enterprise Data and AINot published; tools for nearly all colleaguesAgents only through approved API or MCP channelsNo: coding time only$940k
Bank of America4 of 4CTIO; no Chief AI OfficerNot disclosedControls, no published principlesYes: $250M to $300M in 2026, with a 2027 figure$531k
BBVA4 of 4AI Transformation area at the top of the bankLevel 1, training required before accessEight published principlesResults by process and agent$328k
Berkshire Hathaway1 of 4None at group level; each business decidesNot publishedNot publishedNone published$958k
BlackRock2 of 4No firmwide AI leader named; Aladdin engineering builds the toolsNot publishedGuardrails built into the agent platformNone published$973k
BNY4 of 4Chief Data and AI OfficerLevel 1, near-universal trainingEvery digital employee has a human manager; reviews before releaseSpeed results by use case, no total$417k
Capital One3 of 4Chief Scientist and Chief Enterprise AI OfficerNot publishedEvery agent evaluated continuously, with human oversightNo: declines to quantify benefits$700k
Charles Schwab3 of 4CIO for technology, operations and data; no AI head namedNot publishedGuardrails and human handoffs; outside agents cannot move moneyNo: counts of use cases only$725k
Citigroup4 of 4Group Head of AI, in the COO’s organizationLevel 1, required prompt trainingControls built into the AI platform, human in the loopCapacity and speed results; says benefits exceed costs, no figure$377k
Goldman Sachs3 of 4CIO, sponsored by the top three executivesNone declaredFirmwide AI risk committeeNo: weekly use only$1.23M
JPMorgan Chase4 of 4Chief Data and Analytics Officer, passing to the CTONone declared (use is opt-in)Little publishedEarlier savings estimate, no longer updated$573k
Lloyds Banking Group4 of 4Chief Data and AI OfficerLevel 1, mandatory trainingResponsible AI team and frameworkYes: about £50M in 2025, £100M+ target$377k
Mastercard3 of 4Chief AI and Data OfficerLevel 1, data training required before accessEvery model explainable, governed and monitoredNo: few internal results$824k
Morgan Stanley3 of 4Head of Firmwide AINot published; 98% of adviser teams use the AI assistantEvaluations before every launch; human oversightResults by use case, no total$851k
U.S. Bancorp4 of 4Dedicated AI teams; no single AI leader namedNot published; AI courses offered to allNot publishedResults by use case, no total$417k
Visa4 of 4President of TechnologyNot published; generative AI access for all since 2023Data-use principles built into model monitoringProduct development results, no total$1.32M
Wells Fargo3 of 4Head of AI, reporting to the CEONot published; 90,000+ trainedBusiness heads accountable; rules for customers’ AI agentsNo: coding efficiency only$408k
Klarna
AI Pacesetter
3 of 4CEONot disclosedNone publishedYes: savings and work done by AI$1.24M

Disclosure counts how many of the four strategy questions (where it is going, where it will focus, what it will take and how it will know it is working) a company answers in public. Fluency floor is the lowest level of AI fluency a company expects of all its people, on the five levels of the Lead with AI Fluency Matrix: Avoiding (0), Chatting (1), Collaborating (2), Integrating (3) and Transforming (4). The other columns summarize each company’s profile. Revenue per employee is for each company’s latest fiscal year, in US dollars. Rows marked AI Pacesetter or Notable are companies outside the S&P 100 and STOXX Europe 50. They are shown for contrast and are not counted in any index figure. See all companies in the AI Strategy Index.

Frequently Asked Questions

What is BNY’s AI strategy?

BNY wants AI used by everyone, everywhere, for everything. It treats AI as a capacity generator, starting with its own work, then its products, then new offerings for clients.

‍

Who leads AI at BNY?

CEO Robin Vince sets the direction, and Sarthak Pattanaik is Chief Data and AI Officer. Every digital employee reports to a human manager.

‍

How many BNY employees use AI?

99% of employees had been trained by October 2025. In April 2026, BNY said about half of employees use AI every day and more than half build agents.

‍

What are BNY’s digital employees?

They are autonomous AI workers, more than 140 by August 2026, that work in areas such as payment operations. Each has a persona, its own credentials and a human supervisor.

‍

Is BNY cutting jobs because of AI?

Headcount fell from about 53,400 at the end of 2023 to about 46,600 in September 2026. BNY says AI creates capacity and frees money to invest, and it has not said how much of the decline is due to AI.

‍

How is this profile put together?

Lead with AI profiles use public sources only, such as annual reports, earnings materials, filings and on-record interviews, dated at the time of writing. AI fluency is coded against the Lead with AI Fluency Matrix. Company figures are marked as company-claimed unless they have been independently verified.

About the AI Strategy Index

The Lead with AI AI Strategy Index profiles how the largest companies in the US and Europe approach AI, starting with the S&P 100 and the STOXX Europe 50. Each profile answers four questions: where the company is going with AI, where it will focus, what it will take and how it will know it is working. AI Pacesetters and notable companies outside the two indexes are profiled for contrast and are not counted. Profiles use public sources only, and AI fluency is coded against the Lead with AI Fluency Matrix.

See all company profiles