Verified
U.S. Bancorp
Experimenting- Banking
- 68,520 employees
- Chief AI Officer since mid-2025
- OpenAI, Anthropic, Google models
- AWS and Azure
- No AI return published
U.S. Bancorp deploys AI through an agentic software layer its technology group built and a mix of proprietary, open-source, OpenAI, Anthropic and Google models, with Amazon Bedrock as the base for contact center agents. Chief AI Officer Prashant Mehrotra, hired in mid-2025, leads the program, and engineers lead adoption: 70% to 75% were trained on and using new AI tools within 3 months (Jun 2026). The bank credits AI with part of its productivity gains but has published no dollar return or company-wide user count.
usbank.comTheir approach to AI
- The platform
- U.S. Bancorp's agentic software layer underpins all of its early AI use cases, Chief Information and Technology Officer Dilip Venkatachari said in Jun 2026. The layer is built on an enterprise platform that Chief Digital Officer Dominic Venturo says is 80% reusable.
- The models
- Prashant Mehrotra said in Jul 2026 that U.S. Bancorp uses proprietary and open-source models alongside OpenAI, Anthropic and Google models. Mehrotra said engineers started with GitHub Copilot and now also get tools such as Cognition and Claude Code.
- The reach
- 70% to 75% of U.S. Bancorp's engineers had taken AI tools training and were using the tools within 3 months, Dilip Venkatachari said in Jun 2026. U.S. Bancorp gives no user count for its 68,520 employees.
- The catch
- U.S. Bancorp has published no return it credits to AI. Gunjan Kedia called AI one of several productivity levers on Oct 15, 2026 without giving a figure. U.S. Bancorp's 2025 annual report credits cost savings to digital investment, cloud migration and AI together.
Snapshot
- Employees
- 68,520
- Dec 31, 2025 (2025 Form 10-K)
- Technology spend
- $2.21B (2025)
- Technology and communications expense, up 6.6%; AI spend not broken out
- AI/ML headcount
- Not disclosed
- Prashant Mehrotra cites thousands of builders, engineers, analysts and developers (Jul 2026)
- Who owns AI
- Prashant Mehrotra, Chief AI Officer
- Executive Vice President since mid-2025; Chief Information and Technology Officer Dilip Venkatachari runs the platform and engineering
- Platform
- In-house agentic layer
- Underpins the early AI use cases (Dilip Venkatachari, Jun 2026); Amazon Bedrock for contact center agents
- Reach
- 70-75% of engineers
- Jun 2026; persona-based AI training for about 70,000 employees, no usage count
- Use cases
- Not counted publicly
- Coding, contact centers, developer portal, treasury forecasting, Elavon sales quotes, design review
Key findings
U.S. Bancorp has published no business return it credits to AI. Gunjan Kedia told analysts on Oct 15, 2026 that AI is one, not the only, driver of productivity and that U.S. Bancorp can swap expensive models for cheaper ones. Kedia said the AI program is what could take the bank's efficiency ratio to the low end of its mid to high 50s target, without giving a figure.
U.S. Bancorp's 2025 annual report, published in Mar 2026, credits a generative AI code-review assistant with cutting review time by about 75% and AI tools with saving about 36,000 hours of code-writing time a week. Dilip Venkatachari said in Jun 2026 that 70% to 75% of U.S. Bancorp's engineers were trained on and using new AI tools within 3 months.
Prashant Mehrotra, U.S. Bancorp's Chief AI Officer since mid-2025, said in Jul 2026 that the bank balances proprietary and open-source models with models from OpenAI, Anthropic and Google. Mehrotra said U.S. Bancorp sorts use cases into growth, cost savings and unmeasured productivity, and has never rewarded staff for heavy token use.
U.S. Bancorp said on May 7, 2026 that it will move hundreds of mission-critical applications to AWS and use Amazon Bedrock, Amazon Nova Sonic and Amazon Connect to build AI agents across its lines of business. U.S. Bancorp had named Microsoft Azure its primary cloud in Feb 2022.
Gunjan Kedia said on May 5, 2026 that U.S. Bancorp's workforce likes AI about as much as it liked return-to-office mandates after Covid. Kedia said on Apr 16, 2026 that U.S. Bancorp aims to become an AI-native organization.
Where U.S. Bank sits
- Stage
- Experimenting
U.S. Bancorp sits at Experimenting because it has reported no business return it credits to AI in the last 18 months, for any use case or for the company as a whole. Gunjan Kedia told analysts on Oct 15, 2026 that AI is one, not the only, driver of productivity, and gave no figure. The 2025 savings in U.S. Bancorp's annual report, published in Mar 2026, come from a program labeled AI and automation that runs alongside location, real estate and organizational programs. U.S. Bancorp's AI coding figures, about 36,000 hours of code-writing time saved a week and code-review time down about 75%, carry no cost or headcount equivalent. On rollout, U.S. Bancorp has not said how many of its 68,520 employees use AI tools, although Chief AI Officer Prashant Mehrotra has owned the program since mid-2025. U.S. Bancorp reports persona-based AI training across about 70,000 employees and 70% to 75% of engineers using new AI tools, both in Jun 2026. U.S. Bancorp's AI work on the record since May 2025 includes its developer portal assistant (Oct 2025), Liquidity Manager with Kyriba (Nov 2025) and the AWS agent platform deal (May 2026). U.S. Bancorp falls short of Piloting, which needs one AI-credited return or an officially reported rollout with a user count and a named owner.
How the stages are defined - Ownership
- Chief AI Officer Prashant Mehrotra; technology under Chief Information and Technology Officer Dilip Venkatachari
- Platform
- Hybrid: In-house agentic layer; models and coding tools bought
- Models
- Proprietary and open-source, plus OpenAI, Anthropic and Google
- Hosting
- Microsoft Azure (primary cloud since Feb 2022) and AWS, including Amazon Bedrock and Amazon Connect; where in-house models run is not disclosed
- Reach
- 70% to 75% of engineers (Jun 2026); no company-wide count
How they deploy
From automation to AI-native
Gunjan Kedia told analysts on Apr 16, 2026 that U.S. Bancorp's cost discipline once rested on automation and that going forward the bank is focused on becoming an AI-native organization. U.S. Bancorp organized its productivity work in early 2025 around 4 programs: AI and automation, location optimization, real estate rationalization and organizational simplicity. U.S. Bancorp's 2025 annual report says the programs continue into 2026.
Several models, none locked in
Prashant Mehrotra said in Jul 2026 that U.S. Bancorp uses some proprietary and open-source models and balances them with models from OpenAI, Anthropic and Google so that it is not stuck with one provider. Mehrotra said U.S. Bancorp does this while protecting its most confidential data. Gunjan Kedia said on Oct 15, 2026 that U.S. Bancorp has built the ability to swap out expensive models for cheaper ones as token costs come under scrutiny.
Central start, federated delivery
Prashant Mehrotra said in Jul 2026 that U.S. Bancorp started AI with a centralized organization and is federating it out to the rest of the enterprise. A governance body reviews every use case through a common lens, and U.S. Bancorp pushes units to reuse capabilities instead of rebuilding them. Dominic Venturo said in Jun 2026 that U.S. Bancorp's enterprise platform is 80% reusable and that teams publish what they build back into an internal marketplace.
Measured on outcomes, not tokens
Prashant Mehrotra told American Banker in Jul 2026 that U.S. Bancorp has never been into token maxing, or rewarding heavy token use. Mehrotra said the bank sorts AI ideas into growth drivers, cost savings and productivity that cannot be measured, and focuses on the first 2. Mehrotra said at American Banker's Digital Banking conference in Jun 2026 that the bank picks high-frequency, repeatable work and measures value at every gate from pilot to enterprise scale.
Cautious with customers
Gunjan Kedia said on Oct 15, 2026 that U.S. Bancorp is more measured with client-facing AI applications because of privacy and customer confidence concerns. Kedia said the bank keeps a sharp focus on AI for internal productivity. U.S. Bancorp's May 2026 AWS agreement also includes plans for agentic self-service for customers across voice, chat and text.
- Deployment model
- U.S. Bancorp runs AI through Chief AI Officer Prashant Mehrotra, who sets strategy and execution, and through a technology organization under Chief Information and Technology Officer Dilip Venkatachari. Venkatachari's organization built the agentic software layer behind the early use cases. Dominic Venturo's Chief Digital Office leads digital banking, enterprise platforms, experience design and innovation, and a 2024 job posting for the AI chief role said the role reports to the Chief Digital Officer. Business lines bring use cases forward: corporate functions technology under Andy Bingenheimer runs the contact center platform, the Elavon payments sales force uses AI quotes, and marketing under Michael Lacorazza uses synthetic audiences.
- Model providers
- Prashant Mehrotra said in Jul 2026 that U.S. Bancorp uses proprietary and open-source models alongside models from OpenAI, Anthropic and Google. For engineers, U.S. Bancorp started with GitHub Copilot and now also provides Cognition and Claude Code, and is testing newer tools. U.S. Bancorp's contact center agent assistant runs on Amazon Q in Connect with Anthropic's Claude through Amazon Bedrock, according to a ZenML case study of the bank's AWS presentation. The May 2026 AWS agreement adds Amazon Nova Sonic for voice. Ankit Bhatt declined in Jan 2026 to name the model behind the developer portal assistant.
- Data and infrastructure
- U.S. Bancorp named Microsoft Azure its primary cloud provider for its application portfolio in Feb 2022, with Dilip Venkatachari speaking for the bank. U.S. Bancorp said on May 7, 2026 that it will migrate hundreds of mission-critical applications to AWS in a multiyear program, including payment processing, wealth management and commercial banking systems. Amazon Bedrock and Amazon Connect will be U.S. Bancorp's central base to build and deploy AI agents. U.S. Bancorp moved its contact centers to Amazon Connect from 2019 and finished in 2024, according to the ZenML case study. Gunjan Kedia said in Jan 2026 that preparing for Category II supervision pushed U.S. Bancorp to organize better data over the previous 2 to 3 years. U.S. Bancorp's 2025 annual report says the bank created a data strategy center of expertise linked to its AI and digital teams.
Platforms and internal tools
Agentic software layer2025
U.S. Bancorp's technology group developed agentic software that Dilip Venkatachari said in Jun 2026 underpins all of the bank's early AI use cases inside and outside technology. The layer gives business teams a way to launch new use cases quickly in areas such as mortgages, credit decisioning and know-your-customer checks.
Enterprise platform and reuse marketplace
U.S. Bancorp's enterprise platform lets teams assemble about 80% of a new product from existing tools and processes and publish what they build back into an internal marketplace, Dominic Venturo said in Jun 2026. Prashant Mehrotra's AI playbook builds on the same reuse of archetypes, platforms and patterns.
Code-review assistantBy 2025
U.S. Bancorp's generative AI code-review assistant reviews engineers' code, and the 2025 annual report says it cut review time by about 75%. The CXOTalk episode with Prashant Mehrotra says the results let the bank scale the assistant to thousands of developers.
- Users
- Thousands of developers (Jan 2026)
U.S. Bank Smart Assistant2020
U.S. Bank Smart Assistant is U.S. Bancorp's bilingual customer virtual assistant on the website and mobile app, using machine learning and conversational AI for tasks such as checking balances, tracking credit scores and locking cards.
- Users
- 2.5 million uses a month (Mar 2026)
Contact center platform on Amazon Connect2019, migration finished in 2024
U.S. Bancorp's contact centers run on Amazon Connect, which shows agents a customer's earlier calls, chats, emails and texts so customers need not repeat themselves (Andy Bingenheimer, May 2026). A generative AI assistant on the platform helps agents answer callers faster, and the May 2026 AWS agreement adds agentic self-service across voice, chat and text.
- Models
- Anthropic Claude via Amazon Bedrock (ZenML case study); Amazon Nova Sonic planned
Developer portal assistantOct 2025
U.S. Bancorp's developer portal assistant is a generative AI chat tool that helps business clients and partners find and implement U.S. Bank APIs, explains errors and points to testing scripts and documentation. U.S. Bancorp rolled the assistant out quietly in Oct 2025 and announced it in Jan 2026.
Design AssistantBy Mar 2026
Design Assistant is an in-house U.S. Bancorp AI tool that reviews designers' work, flags likely issues and suggests improvements across digital products, built after a mid-2025 review of design workflows.
U.S. Bank Liquidity ManagerNov 18, 2025
U.S. Bank Liquidity Manager, powered by Kyriba, gives mid-sized and large corporate clients cash forecasting with Cash AI, which predicts future cash flows from historical data. Liquidity Manager also offers cash positioning, pooling and multi-bank reporting inside the SinglePoint treasury platform.
Vendors and partners
Amazon Web Services2019 (Amazon Connect); expanded May 2026
Cloud for hundreds of mission-critical applications; Amazon Bedrock, Nova Sonic and Connect for AI agents and contact centers · U.S. Bancorp also uses AWS Transform and Amazon Kiro to speed up its migration and AWS training programs to upskill technologists (May 2026).
MicrosoftFeb 2022
Microsoft Azure, primary cloud provider for the application portfolio
OpenAI
Model provider · Prashant Mehrotra named OpenAI among U.S. Bancorp's model sources in Jul 2026, without saying which products are used.
Anthropic
Model provider; Claude Code for engineers; Claude in the contact center assistant · The Claude contact center detail comes from a ZenML case study, not from U.S. Bancorp.
Google
Model provider · Prashant Mehrotra named Google among U.S. Bancorp's model sources in Jul 2026.
GitHub (Microsoft)
GitHub Copilot, the first coding assistant U.S. Bancorp gave its engineers
Cognition
AI coding tools for engineers (Jul 2026)
KyribaNov 2025
Cash AI forecasting engine behind U.S. Bank Liquidity Manager
MastercardSep 2025
Agent Pay for AI-agent purchases · Mastercard said on Sep 10, 2025 that Citi and U.S. Bank Mastercard cardholders would be the first to try Agent Pay.
Supernatural AI2024
Synthetic audience avatars for marketing research
Leadership
AI, data and platform leaders
Executives
Board
AI org map
Chief AI Officer setting strategy, with delivery federating into technology, digital and the businesses
U.S. Bancorp hired Prashant Mehrotra as Chief AI Officer in mid-2025 to lead AI strategy and execution. Mehrotra said in Jul 2026 that the bank started with a centralized AI organization and is federating it out. A 2024 U.S. Bank job posting for the Chief AI/ML Officer role said the role reports to the Chief Digital Officer, Dominic Venturo. U.S. Bancorp has not published Mehrotra's reporting line. Chief Information and Technology Officer Dilip Venkatachari and Venturo sit on the managing committee. U.S. Bancorp's board added a Technology Committee on Jan 1, 2026 that reviews the bank's AI adoption strategy.
U.S. Bancorp's AI work sits in the Chief AI Office, the technology organization and the Chief Digital Office. The Chief AI Office under Prashant Mehrotra began as a centralized organization and is federating out to the businesses. Engineers in the technology organization under Dilip Venkatachari built the agentic software layer and lead adoption of AI coding tools. The Chief Digital Office under Dominic Venturo owns digital banking, enterprise platforms and experience design. U.S. Bancorp has not published how many AI or machine learning specialists it employs, and Mehrotra speaks of thousands of builders, software engineers, analysts and developers. U.S. Bancorp's 2026 proxy refers to an Artificial Intelligence and Machine Learning Center of Excellence that has briefed the board since 2024.
Reporting lines
Chairman and CEO
Technology
Dilip Venkatachari, Chief Information and Technology Officer
Managing committee member; line to the CEO inferred, not published
Information Security Services
Co-CISOs
2025 Form 10-K
Corporate Functions Technology
Andy Bingenheimer, CIO
Reporting line inferred from the title, not published
Chief Digital Office
Dominic Venturo, Chief Digital Officer
Digital banking, enterprise platforms, experience design, innovation and Digital Assets and Money Movement; line to the CEO inferred
Chief AI Office
Prashant Mehrotra, Chief AI Officer
Line taken from the 2024 job posting for the role; not confirmed
Operations
Toby Clements, COO
More than 16,000 employees in client service centers and global operations (Mar 2026)
Risk Management
Troy Remington, interim Chief Risk Officer
The Chief Risk Officer chairs the Executive Risk Committee; Jodi Richard left in Oct 2026 and a successor search is under way
U.S. Bancorp has not published where Prashant Mehrotra reports. The Chief AI/ML Officer job posting from 2024 placed the role under the Chief Digital Officer. The 2025 Form 10-K lists Dominic Venturo as a managing committee member who is not deemed an executive officer, and lists Dilip Venkatachari among the executive officers.
Named people
Use cases
U.S. Bancorp has not published a count of its AI use cases. The use cases U.S. Bancorp has described in production include AI coding tools and generative AI assistants for code review and for contact center agents. U.S. Bancorp has also described the Smart Assistant for customers, a developer portal assistant, AI cash forecasting with Kyriba, AI quotes for Elavon's sales force and a Design Assistant. U.S. Bancorp reports no business return it credits to AI for any of them: its coding figures are hours saved and review time cut, with no cost or headcount equivalent. At company level, Gunjan Kedia said on Oct 15, 2026 that AI contributes to productivity without giving a figure. U.S. Bancorp's 2025 savings sit in a program it labels AI and automation.
| Use case | Area | Status | ROI |
|---|---|---|---|
| Code-review assistant U.S. Bancorp's generative AI code-review assistant cut review time by about 75%. The 2025 annual report credits AI tools with about 36,000 hours of code-writing time saved each week, with no cost or headcount equivalent. The CXOTalk episode with Prashant Mehrotra says the results beat expectations by more than 50% and let the bank scale the assistant to thousands of developers. | Technology | In production | |
| AI coding assistants and agents U.S. Bancorp engineers started with GitHub Copilot and now also use Cognition and Claude Code to write and test more code. Prashant Mehrotra said U.S. Bancorp tracks cycle time and production incidents. Dilip Venkatachari said in Jun 2026 that 70% to 75% of engineers were trained and actively using new tools within 3 months. | Technology | In production | |
| Contact center agent assistant A generative AI assistant helps U.S. Bancorp's contact center agents answer callers faster. A ZenML case study of the bank's AWS presentation describes Amazon Q in Connect transcribing calls, detecting intents and suggesting knowledge base answers drawn by Claude, first as a production pilot for some business lines. The CXOTalk episode with Prashant Mehrotra says response times fell from minutes to tens of seconds. | Operations | In production | |
| Agentic self-service for customers U.S. Bancorp plans agentic self-service powered by generative AI across voice, chat and text on Amazon Connect, Amazon Bedrock and Amazon Nova Sonic, with a choice between self-service and a human at any hour. | Operations | Announced | |
| U.S. Bank Smart Assistant U.S. Bank Smart Assistant is a bilingual virtual assistant on the website and mobile app that uses machine learning and conversational AI to check balances, track credit scores and lock cards. Customers used Smart Assistant 2.5 million times a month (Mar 2026). | Consumer banking | In production | |
| AI next-best suggestions in branches Gunjan Kedia said in Jan 2026 that AI-powered next best solutions and other tools given to branch staff during 2025 helped branch performance turn up. | Consumer banking | In production | |
| Developer portal assistant U.S. Bancorp's generative AI assistant helps the business clients and partners who use its developer portal find and implement APIs, explains errors and promotes practices such as account tokenization. Ankit Bhatt said the assistant should cut average API implementation time by weeks. | Embedded and corporate banking | In production | |
| Liquidity Manager cash forecasting U.S. Bank Liquidity Manager, powered by Kyriba, uses Cash AI to predict corporate clients' future cash flows from historical data, alongside automated cash positioning, pooling and multi-bank reporting. | Treasury management | In production | |
| AI merchant quotes and proposals U.S. Bancorp's Elavon sales force uses generative AI, language models, vision models and predictive analytics to read a merchant's current spending and produce a comparable quote and proposal in near real time instead of weeks. Prashant Mehrotra said the bank measures conversion and onboarding speed but gave no figure. | Payments (Elavon) | In production | |
| Design Assistant Design Assistant reviews U.S. Bank designers' work against performance data, flags where problems usually surface and suggests fixes before handoff to engineering. | Digital | In production | |
| Synthetic audiences for campaign research U.S. Bancorp's marketing team uses Supernatural AI avatars of its target customers as synthetic focus groups to develop campaign strategy and test creative. Michael Lacorazza said tests with humans showed a 90% to 95% overlap in themes. | Marketing | In production | |
| AI in building and testing controls Prashant Mehrotra said in Jul 2026 that U.S. Bancorp is looking at AI in risk processes to build and test controls more efficiently, with guardrails so that fixing the output does not take longer than doing the work by hand. | Risk | Pilot | |
| Fraud detection and compliance automation U.S. Bancorp said in May 2026 that it is pursuing generative AI use cases in fraud detection, compliance automation, developer productivity and customer experience with AWS. | Enterprise | Announced |
Adoption
U.S. Bancorp's AI adoption is clearest among engineers. Dilip Venkatachari said in Jun 2026 that 70% to 75% of engineers had completed training on new AI tools and were actively using them within 3 months, and the code-review assistant reached thousands of developers. For the wider workforce, U.S. Bancorp reports training, not usage. An AI Essentials Channel in its Skills Academy (2025) teaches staff to use company-approved AI tools, and Prashant Mehrotra said in Jun 2026 that persona-based AI training has rolled out across about 70,000 employees. U.S. Bancorp has not said how many employees use AI tools each week. Gunjan Kedia said on May 5, 2026 that most workers remain skeptical of AI and that confident users will need to show others what works.
By the numbers
Rollout, first pilot to today
Nov 2019
U.S. Bancorp hired Tanushree Luke from Capital One as head of AI at U.S. Bank, to bring AI into product development, innovation and strategic planning.
2020
U.S. Bank rolled out a voice assistant in its mobile app, and added Spanish in 2022.
Feb 2022
U.S. Bancorp named Microsoft Azure its primary cloud provider for its application portfolio.
2024
U.S. Bank advertised for a Chief AI/ML Officer to set strategy for its AI and machine learning portfolio, reporting to the Chief Digital Officer.
Early 2025
U.S. Bancorp organized its productivity work around 4 programs, among them AI and automation.
Mid-2025
Prashant Mehrotra became U.S. Bancorp's Chief AI Officer, after heading the AI Center of Excellence at Allstate.
Oct 2025
U.S. Bank quietly rolled out a generative AI assistant on its developer portal for business clients and partners.
Oct 2025
U.S. Bancorp's board created a Technology Committee, effective Jan 1, 2026, to oversee technology strategy and review the company's AI adoption strategy.
Nov 18, 2025
U.S. Bank launched Liquidity Manager with Kyriba, an AI cash forecasting tool for corporate treasurers.
Mar 2026
U.S. Bancorp's 2025 annual report said its code-review assistant cut review time about 75%, AI tools save about 36,000 hours of code-writing time a week, and a generative AI assistant supports contact center agents.
May 2026
U.S. Bancorp agreed to move hundreds of mission-critical applications to AWS and build AI agents across its lines of business on Amazon Bedrock and Amazon Connect.
Jun 2026
Dilip Venkatachari said 70% to 75% of U.S. Bancorp's engineers were trained on and using new AI tools within 3 months, and Prashant Mehrotra said persona-based AI training had reached about 70,000 employees.
Oct 2026
Gunjan Kedia said U.S. Bancorp remains focused on the AI program and expects it to keep contributing to productivity, while staying measured on client-facing AI.
Economics & controls
Spend and ROI
Technology budget
U.S. Bancorp's technology and communications expense was $2.21 billion in 2025, up 6.6% from $2.07 billion in 2024, which the bank attributed to infrastructure and technology development. U.S. Bancorp said it invested more than $2.5 billion in 2025 to drive organic growth while keeping total expenses essentially flat. U.S. Bancorp does not break out what it spends on AI.
What AI itself costs
U.S. Bancorp has not disclosed its AI spend or token costs. Prashant Mehrotra said in Jul 2026 that model providers have raised prices to match value and that the bank has never rewarded heavy token use. Gunjan Kedia said on Oct 15, 2026 that the debate over whether token costs are overtaking returns is relevant and that U.S. Bancorp can swap expensive models for cheaper ones.
Reported ROI
U.S. Bancorp reports no business return it credits to AI alone. The 2025 annual report says the cloud migration and rapid deployment of AI, on top of earlier digital investment, helped achieve meaningful cost savings in core operations and keep expenses essentially flat. The report books those savings to a program labeled AI and automation, alongside location, real estate and organizational programs, with no AI share. Gunjan Kedia said on Oct 15, 2026 that productivity initiatives, including the AI programs, improved the efficiency ratio. When an analyst asked for numbers, Kedia called AI one, not the only, reason for good productivity without giving a figure.
Productivity markers
U.S. Bancorp's generative AI code-review assistant cut review time by about 75%, and AI tools save about 36,000 hours of code-writing time a week (2025 annual report, Mar 2026). U.S. Bancorp gives no cost or headcount equivalent. Prashant Mehrotra said in Jul 2026 that AI cut Elavon merchant quotes from weeks to near real time with a real dollar impact on conversion, without a figure. Michael Lacorazza said on a podcast recorded at CES 2026 that AI personalization generates tens of millions of dollars of value, according to the host's summary. U.S. Bancorp has not published that figure.
Headcount
U.S. Bancorp employed 68,520 people at Dec 31, 2025, and its press releases put the workforce at about 70,000. U.S. Bancorp has not tied any change in headcount to AI. Gunjan Kedia said in Jan 2026 that the AI boost in big expense pools adds to the productivity from 6 years of digital investment. Kedia said in Oct 2026 that the AI program could accelerate and begin to affect the productivity of the whole firm.
Governance and risk
Board oversight
U.S. Bancorp's board created a Technology Committee in Oct 2025, effective Jan 1, 2026, to oversee technology strategy, modernization and cybersecurity technology and to review the company's strategy for adopting AI. Yusuf Mehdi of Microsoft chairs the Technology Committee, which meets jointly with the Risk Management and Audit Committees from 2026. The full board received reports during 2025 on U.S. Bancorp's adoption and governance of AI.
Risk partners from the start
U.S. Bancorp brings risk, cybersecurity and other functions into AI projects at the start, and Prashant Mehrotra said in Jan 2026 that the approach cut governance approval times in half over 6 months. Mehrotra calls the principle responsible by design, with compliance built into AI products, and a governance body reviews every use case through a common lens.
Model risk
U.S. Bancorp's 2025 annual report says the bank uses several models built on AI or machine learning that bring risks of algorithmic bias, data needs and explainability. The report says flawed models could mislead business decisions, customers or regulators. The Executive Risk Committee, chaired by the Chief Risk Officer, oversees U.S. Bancorp's risk management framework. The Cybersecurity and Technology Governance Committee, co-chaired by the co-CISOs, the Chief Technology Risk Officer and the Head of Enterprise Architecture, governs technology risk.
Data and approved tools
Prashant Mehrotra said in Jul 2026 that U.S. Bancorp wants to protect its most confidential data when using outside models. Mehrotra said each engineered AI solution defines the model, the subset of data it can act on and the intended outcome. U.S. Bancorp's AI Essentials Channel trains staff to use company-approved AI tools, and Ankit Bhatt said the developer portal assistant is gated to the bank's own documentation.
Disclosed risks
U.S. Bancorp's 2025 annual report warns that AI lets attackers automate large-scale attacks and social engineering. The report also warns that the bank's AI and machine learning could produce biased or inaccurate decisions and harm its brand, and that the lack of a uniform AI regulatory framework could raise compliance costs. Dilip Venkatachari said in Jun 2026 that inbound threats will soon outpace human defenders and that the answer is AI and automation on the defense side.
Recent signals
Oct 29, 2026
Banking Dive reported that U.S. Bancorp's Vice Chair and Chief Risk Officer Jodi Richard is leaving to become chief risk officer of American Express, with chief credit officer Troy Remington as interim risk chief.
Oct 15, 2026
Gunjan Kedia told analysts that AI is one, not the only, driver of U.S. Bancorp's productivity and that the bank can swap expensive models for cheaper ones. Kedia said U.S. Bancorp is measured on client-facing AI and has seen no third-party AI agents moving deposits.
Sep 2, 2026
American Banker set out Prashant Mehrotra's 6-step AI playbook from the Digital Banking conference, including persona-based AI training across U.S. Bancorp's 70,000 employees.
Jul 30, 2026
Prashant Mehrotra told American Banker that U.S. Bancorp uses proprietary, open-source, OpenAI, Anthropic and Google models, gives engineers Cognition and Claude Code, and has never rewarded heavy token use.
Jun 1, 2026
American Banker named Dilip Venkatachari and Dominic Venturo among its Most Innovative People in Finance; Venkatachari said 70% to 75% of U.S. Bancorp's engineers were using new AI tools within 3 months.
May 7, 2026
U.S. Bancorp expanded its AWS agreement to migrate hundreds of mission-critical applications and build AI agents on Amazon Bedrock, Amazon Nova Sonic and Amazon Connect. Andy Bingenheimer described the contact center platform at an AWS event the same day.
May 5, 2026
Gunjan Kedia said at the WSJ Future of Everything event that U.S. Bancorp's workforce likes AI about as much as it liked return-to-office mandates after Covid.
Apr 16, 2026
Gunjan Kedia said on the Q1 2026 call that U.S. Bancorp aims to become an AI-native organization, and John Stern said AI lets the bank simplify fund services and corporate trust operations.
Mar 31, 2026
PYMNTS reported that U.S. Bank introduced Design Assistant, an in-house AI tool that reviews designers' work and flags issues before release.
Mar 26, 2026
U.S. Bancorp named Toby Clements chief operations officer from Apr 13, succeeding Souheil Badran, who led the push to bring automation and AI to the contact centers.
Mar 10, 2026
U.S. Bancorp's 2025 annual report said a generative AI code-review assistant cut review time about 75% and AI tools save about 36,000 hours of code-writing time a week. U.S. Bancorp's proxy, filed the same day, described the new board Technology Committee.
Jan 20, 2026
Gunjan Kedia told analysts that U.S. Bancorp is seeing an AI boost in big expense pools on top of 6 years of digital investment.
Jan 15, 2026
U.S. Bank announced a generative AI assistant on its developer portal that Ankit Bhatt said should cut client API implementation time by weeks.
Jan 2026
Prashant Mehrotra said on CXOTalk that bringing risk partners in early cut U.S. Bancorp's AI governance approval times in half over 6 months.
Nov 18, 2025
U.S. Bank launched Liquidity Manager, powered by Kyriba, with AI cash forecasting for corporate treasurers.
Quick answers
- How does U.S. Bancorp deploy AI?
- U.S. Bancorp deploys AI through an agentic software layer its technology group built and a mix of proprietary, open-source, OpenAI, Anthropic and Google models, with Amazon Bedrock as the base for contact center agents. Chief AI Officer Prashant Mehrotra, hired in mid-2025, leads the program, and engineers lead adoption: 70% to 75% were trained on and using new AI tools within 3 months (Jun 2026). The bank credits AI with part of its productivity gains but has published no dollar return or company-wide user count.
- What does U.S. Bancorp use AI for?
- U.S. Bancorp has not published a count of its AI use cases. The use cases U.S. Bancorp has described in production include AI coding tools and generative AI assistants for code review and for contact center agents. U.S. Bancorp has also described the Smart Assistant for customers, a developer portal assistant, AI cash forecasting with Kyriba, AI quotes for Elavon's sales force and a Design Assistant. U.S. Bancorp reports no business return it credits to AI for any of them: its coding figures are hours saved and review time cut, with no cost or headcount equivalent. At company level, Gunjan Kedia said on Oct 15, 2026 that AI contributes to productivity without giving a figure. U.S. Bancorp's 2025 savings sit in a program it labels AI and automation.
- Who leads AI at U.S. Bancorp?
- AI at U.S. Bancorp is led by Prashant Mehrotra (EVP and Chief AI Officer); Dilip Venkatachari (Senior EVP and Chief Information and Technology Officer); Dominic Venturo (Senior EVP and Chief Digital Officer). Ownership of AI on the org chart: Chief AI Officer Prashant Mehrotra; technology under Chief Information and Technology Officer Dilip Venkatachari.
- What AI platform does U.S. Bancorp use?
- U.S. Bancorp's agentic software layer underpins all of its early AI use cases, Chief Information and Technology Officer Dilip Venkatachari said in Jun 2026. The layer is built on an enterprise platform that Chief Digital Officer Dominic Venturo says is 80% reusable.
- Which AI models does U.S. Bancorp use?
- Prashant Mehrotra said in Jul 2026 that U.S. Bancorp uses proprietary and open-source models alongside OpenAI, Anthropic and Google models. Mehrotra said engineers started with GitHub Copilot and now also get tools such as Cognition and Claude Code.
- How many U.S. Bank employees use AI?
- 70% to 75% of U.S. Bancorp's engineers had taken AI tools training and were using the tools within 3 months, Dilip Venkatachari said in Jun 2026. U.S. Bancorp gives no user count for its 68,520 employees.
- What return has U.S. Bancorp reported from AI?
- U.S. Bancorp sits at Experimenting because it has reported no business return it credits to AI in the last 18 months, for any use case or for the company as a whole. Gunjan Kedia told analysts on Oct 15, 2026 that AI is one, not the only, driver of productivity, and gave no figure. The 2025 savings in U.S. Bancorp's annual report, published in Mar 2026, come from a program labeled AI and automation that runs alongside location, real estate and organizational programs. U.S. Bancorp's AI coding figures, about 36,000 hours of code-writing time saved a week and code-review time down about 75%, carry no cost or headcount equivalent. On rollout, U.S. Bancorp has not said how many of its 68,520 employees use AI tools, although Chief AI Officer Prashant Mehrotra has owned the program since mid-2025. U.S. Bancorp reports persona-based AI training across about 70,000 employees and 70% to 75% of engineers using new AI tools, both in Jun 2026. U.S. Bancorp's AI work on the record since May 2025 includes its developer portal assistant (Oct 2025), Liquidity Manager with Kyriba (Nov 2025) and the AWS agent platform deal (May 2026). U.S. Bancorp falls short of Piloting, which needs one AI-credited return or an officially reported rollout with a user count and a named owner.
Open questions
- How many of U.S. Bancorp's 68,520 employees use AI tools each week, and which general-purpose assistant, if any, do non-technical staff use?
- What share of the 2025 cost savings from the AI and automation program does U.S. Bancorp credit to AI, in dollars or headcount?
- What do the 36,000 hours of code-writing time saved each week translate to in cost, capacity or output per engineer?
- What returns do the contact center agent assistant, Elavon's AI quotes, the developer portal assistant and Liquidity Manager produce, given that U.S. Bancorp has published none?
- Where does Prashant Mehrotra report, and how large is the Chief AI Office and the AI and Machine Learning Center of Excellence?
- How is model traffic split between OpenAI, Anthropic, Google, open-source and proprietary models, and which cloud runs the in-house models?
- How much does U.S. Bancorp spend on AI, including token costs, within its $2.21 billion of 2025 technology and communications expense?
- When will U.S. Bancorp launch agentic self-service to customers, given Gunjan Kedia's caution on client-facing AI in Oct 2026?
- Did U.S. Bancorp's synthetic audiences or AI personalization produce a measured revenue gain, as Chief Marketing Officer Michael Lacorazza has suggested?
Sources
U.S. Bancorp
- U.S. Bank launches AI-driven cash forecasting tool with Kyriba (Nov 18, 2025)
- Three U.S. Bank leaders named to American Banker's Most Innovative list (Jun 2026)
- Souheil Badran to retire as chief operations officer (Jan 29, 2026)
- Toby Clements to become chief operations officer (Mar 26, 2026)
- Job description: Chief AI/ML Officer, U.S. Bank (2024), via InnoLead
Filings, letters and investor materials
- U.S. Bancorp 2025 Annual Report (Mar 2026)
- U.S. Bancorp 2025 Form 10-K (filed Feb 23, 2026)
- U.S. Bancorp 2026 Proxy Statement (filed Mar 10, 2026)
- Q3 2026 earnings call transcript, The Motley Fool (Oct 15, 2026)
- Q1 2026 earnings call transcript, The Motley Fool (Apr 16, 2026)
- Q4 2025 earnings call transcript, The Motley Fool (Jan 20, 2026)
Partners
News
- American Banker: Beyond token maxing, how U.S. Bank's AI chief navigates costs (Jul 30, 2026)
- American Banker: Benchmarking AI adoption, what U.S. Bank's playbook tells us (Sep 2, 2026)
- American Banker: Dilip Venkatachari, #15 Most Innovative People in Finance (Jun 1, 2026)
- American Banker: Dominic Venturo, #41 Most Innovative People in Finance (Jun 1, 2026)
- American Banker: US Bank, CoBank, Rocket share AI use cases at AWS event (May 7, 2026)
- American Banker: U.S. Bank gives embedded banking clients a gen AI assistant (Jan 15, 2026)
- American Banker: U.S. Bancorp recruits Capital One exec to be its AI chief (Nov 6, 2019)
- American Banker Digital Banking: Prashant Mehrotra speaker profile (2026)
- CXOTalk: U.S. Bank's Chief AI Officer on strategy, governance and scaling AI (Jan 2026)
- Show notes of the CXOTalk episode, reposted (Jan 2026)
- Dow Jones via Tiger Brokers: US Bank CEO says employees dislike AI as much as return-to-office mandates (May 5, 2026)
- CIO Dive: U.S. Bank shifts critical apps to AWS in AI push (May 2026)
- PYMNTS: BofA and U.S. Bank turn AI loose on internal bottlenecks (Mar 31, 2026)
- PYMNTS: U.S. Bancorp says AI agents are not moving consumer deposits yet (Oct 2026)
- Retail Banker International: US Bank taps Microsoft (Feb 23, 2022)
- Marketing Dive: U.S. Bank's Power of Us campaign with Supernatural AI (Jun 3, 2024)
- AdExchanger: US Bank is using synthetic audiences to generate real customer insights
- Speed of Culture podcast summary: how U.S. Bank uses AI to drive revenue, Michael Lacorazza at CES 2026
- Banking Dive: U.S. Bank scouts for agentic AI at CES (Jan 2025)
- Banking Dive: U.S. Bank risk chief Jodi Richard departs for American Express (Oct 29, 2026)
Data confidence
- Confirmed
- Stated in a primary or reputable public source.
- Estimated
- Derived or approximated from public reporting.
- Unconfirmed
- Reported, but not yet corroborated.
U.S. Bancorp profile · Enterprise AI diffusion report by Imagine · facts checked against their sources on Nov 9, 2026
