Capital One AI Strategy (2026): Use Cases, Leaders

Build, don't buy: a research lab, a chief scientist from Amazon, and agents that act — starting with car buying.

How is Capital One using AI?

Capital One set out to be 'a bank that a technology company would build' and its AI strategy follows: it hires scientists rather than buying tools, runs an Applied AI research group on model training and multi-agent reasoning, funds university AI research at UVA, Illinois and through the National Science Foundation, and holds the second-largest AI talent pool in banking after JPMorgan, per Evident, which has ranked it second overall three times running. Its first agentic product, Chat Concierge, reached dealers' websites in March 2025; its chief scientist, Prem Natarajan, came from running Amazon's Alexa AI. Discover (acquired May 2025) added scale and, per Evident, more than 2,200 AI staff; Brex (closed April 2026) added an agentic business-payments suite; and founder-CEO Richard Fairbank frames the moment as 'the greatest change in human history' in which 'the threat of obsolescence is the greatest ever'.

At a glance

Lead bank

Capital One, N.A.

Assets (lead bank, Mar 31, 2026)

$672 billion

Charter

National bank (OCC)

Headquarters

McLean, VA

Flagship platform

Chat Concierge and the in-house agent stack

AI leadership

Richard Fairbank · Prem Natarajan · Andy Ozment · Robert M. Alexander · Amy Lenander · Milind Naphade · Chris Nims · Sanjiv Yajnik · Rashmi Shetty

Evident's AI index: second among the 50 largest banks, behind JPMorgan, in three consecutive editions; second-largest AI talent pool after adding 2,200 AI professionals, largely through the Discover acquisition completed May 2025.
Chief Scientist and Head of Enterprise AI Prem Natarajan, formerly VP of Alexa AI at Amazon; the bank says it builds its AI stack and treats AI as a scientific discipline.
Chat Concierge (March 2025): the first proprietary multi-agent assistant, on dealer websites through the Auto Navigator platform; dealers using it reported a 55% improvement in engagement and serious-lead metrics.
Research investments: Applied AI research on LLM lifecycle and agentic reasoning; $4.5 million UVA initiative (October 2025); NSF partnership at $1 million a year for five years; agentic-AI research awards at Illinois; an inaugural AI Symposium (April 2026).
Brex acquired for $5.15 billion (announced January 2026, closed April 2026): 35,000 clients and a suite of agentic AI products for business payments.
Capital One Software sells enterprise data-security tooling (Databolt) aimed at 'secure AI adoption' — the bank as vendor.
Eno, the bank's customer assistant since 2017, began as a text-message chatbot with natural-language processing built in-house; it now works in the mobile app, by text and email, through notifications (including on smart watches) and online from a computer, where it answers account questions and flags charges that might surprise customers.
Proprietary models: the bank says its generative AI models are built by customising open-source LLMs with synthetic and open-source data (a training set of more than 20 trillion tokens), and its MACAW multi-agent workflow — understanding, reasoning, validation and explanation agents on fine-tuned open-weight models — summarises fraud calls and also runs Chat Concierge.
Intellectual property: more than 5,000 granted U.S. patents (June 2025), and 472 AI patents granted in 2024, sixth overall and the highest-ranked bank on Harrity's 2025 AI Patent 100 list; the 2025 10-K warns that AI output 'may not be subject to copyright or other intellectual property protection'.
Dealer AI: dealers using Chat Concierge reported a 55% improvement in engagement and serious-lead metrics, per AI Foundations head Milind Naphade; ProtectID, an AI fraud screen inside Dealer Navigator, won a 2026 FinTech Breakthrough award.
Earnings calls, October 2025 to July 2026: Fairbank tells investors the bank's tech stack was 'built from the outset working backwards from the AI revolution', that transforming banking requires AI 'deeply embedded' in operations and risk management, and that it keeps investing in 'AI infrastructure and specific AI experiences'.
Responsible AI: an AI governance and model-risk disclosure in the 10-K, published research on guardian models and LLM alignment, the ASKS Center for generative-AI safety at Illinois, the open-sourced VulnHunter code-security agent (July 2026) and co-authorship of six banks' agentic-commerce principles (September 2026).

The platform

What is Chat Concierge and the in-house agent stack?

Chat Concierge, launched in March 2025, was the bank's first proprietary multi-agent conversational assistant, embedded on car dealers' websites to compare vehicles, schedule test drives and hand qualified buyers to sales; the wider stack is built by an Applied AI research team working on LLM training, customisation and multi-agent reasoning.[7][8][17]

Proprietary AI

What proprietary AI solutions has Capital One built?

Capital One builds most of its AI itself. Its generative AI models are developed by customising open-source large language models with synthetic and publicly available data, from a training set of more than 20 trillion tokens; Chat Concierge, its first proprietary multi-agent assistant, runs on Meta's Llama customised with Capital One data, and its MACAW multi-agent workflow chains fine-tuned open-weight models in fraud-call servicing. Around them sit the Eno assistant (2017), ProtectID fraud models for car dealers, open-source tools such as Data Profiler, rubicon-ml and VulnHunter, and Databolt, which it sells to other companies. The bank held more than 5,000 granted U.S. patents by June 2025 and was granted 472 AI patents in 2024.[32][7][39][10]

Solution	What it does	Since
In-house generative AI models	Open-source large language models customised with synthetic and publicly available data; the training datasets exceed 20 trillion tokens	Disclosure retrieved Sep 2026[32]
Chat Concierge	Multi-agent assistant on car dealers' websites that answers questions, compares vehicles and schedules test drives; built on Meta's open-source Llama customised with Capital One's proprietary data	March 2025[7]
MACAW multi-agent workflow	Chains understanding, reasoning, validation and explanation agents on fine-tuned open-weight models, first in fraud-call servicing; also behind Chat Concierge (described in sponsored content)	Described Aug 2026[39]
Eno	Customer assistant in the app, by text and email; its natural-language processing was built in-house	Unveiled March 2017[6][4][3]
ProtectID	Proprietary fraud-detection models that give auto dealers a real-time risk assessment inside Capital One's dealer platform	Award announced March 2026[36]
VulnHunter	Agentic AI tool, built internally, that finds security flaws in code; released as open source	July 2026[38]
Data Profiler and rubicon-ml	Open-source Python tools for data analysis, monitoring and sensitive-data detection (Data Profiler) and for capturing model-development runs (rubicon-ml)	February 2021[31]
Databolt	Tokenization product from Capital One Software that protects sensitive data, extended in March 2026 to unstructured data used by AI	Expanded March 2026[42]

Timeline

What has Capital One done on AI, and when?

Figure 1 · Capital One's AI moves by year, from the timeline below. Before 2024 is collapsed into one column. Filled boxes are the current year. Every entry cites its source in the list at the end of the page.

Sep 29, 2026
Natarajan: cutting latency and token costs with smaller, open-weight models[46]
Chief enterprise AI officer Prem Natarajan says the bank has put more emphasis on agentic AI infrastructure to reduce latency and lower token costs, building customised open-weight and specialised models with less context targeted at the desired output; customer-service agents now edit only one-fifth of AI-drafted copy.
Sep 28, 2026
Shetty: 'agentic AI now must be treated as an end-to-end system'[45]
VP of enterprise AI Rashmi Shetty says 14 years of data-infrastructure work gives agents a strong foundation and context, and that the bank adopts platform-first governance — rule-based validation, sandbox simulations, observability of agent trajectories, standardised harnesses and human-in-the-loop controls for high-risk actions — before developers build agents, rather than retrofitting it later.
Sep 22, 2026
Agentic-commerce principles with five other banks[41]
With Bank of America, CBA, ING, NatWest and ASB, a paper on transparency, safety, privacy and data, choice and interoperability.
Aug 13, 2026
MACAW and open-weight models, described[39]
A machine-learning engineering leader explains the multi-agent fraud-call workflow and fine-tuning open-weight models on proprietary data.
Aug 3, 2026
AI fluency program[14]
'We don't just use AI—we build AI'; sustaining an edge requires more than off-the-shelf models.
Jul 21, 2026
Q2 2026 call: AI infrastructure investment continues; Brex models to be calibrated[28][29]
Investment focus stays on the technology transformation and AI infrastructure; full Brex acceleration needs data pipelines and model calibration.
Jul 16, 2026
VulnHunter open-sourced[38]
An internally built agentic AI code-security tool released on GitHub under Apache 2.0; the CISO cites the brief window before AI attack tools spread.
Apr 23, 2026
Inaugural Capital One AI Symposium[13]
Two days in McLean on multi-sector AI partnerships.
Apr 21, 2026
Q1 2026 call: AI leverage is 'vastly greater' when embedded[27]
Fairbank: 14th year of the technology transformation; continued investment in AI infrastructure and specific AI experiences.
Apr 20, 2026
ICLR 2026: alignment and guardian-model research[34]
Gold sponsor and contributor on LLM alignment, uncertainty and responsible agentic systems, including the DynaGuard guardian model.
Apr 7, 2026
Brex deal closes[19][30]
35,000 clients and a suite of agentic AI products join the business-banking strategy.
Mar 23, 2026
Databolt expands for secure AI adoption[42]
Capital One Software's data-security product adds capabilities for sensitive unstructured data.
Mar 19, 2026
ProtectID wins a fraud-prevention award[36]
The AI-driven dealer fraud screen inside Dealer Navigator is named a 2026 FinTech Breakthrough winner.
Mar 17, 2026
'Building AI moats': the chief scientist on Bloomberg[22]
Why the bank builds rather than buys its AI stack; AI as a capacity multiplier.
Feb 19, 2026
2025 10-K: 'our evolving use of AI' is a risk factor[24]
Models and AI in pricing, fraud detection, loan grading, credit and software development; agentic-commerce fraud liability; uncertain IP protection for AI output.
Jan 22, 2026
Q4 2025 call: a stack built 'working backwards from the AI revolution'[26]
Fairbank says the bank is 'now building AI solutions across our businesses' and that Brex has deployed in-house AI agents for expense management and audit.
Jan 22, 2026
Agrees to buy Brex for $5.15 billion[23]
Corporate cards, expense software and nearly $13 billion in deposits at partner banks.
Nov 20, 2025
Applied AI research agenda published[8]
LLM training, customisation and alignment; synthetic data; multi-agent reasoning.
Oct 27, 2025
$4.5 million UVA AI research initiative[12]
Including the Capital One AI Research Neighborhood at UVA Engineering.
Oct 21, 2025
Q3 2025 call: AI 'deeply embedded', not bolted on[25]
Fairbank: AI-driven experiences are 'a new front' in premium cards; the opportunities require 'significant investment in AI and AI talent'.
Oct 8, 2025
Second on the Evident AI index, again[18]
'A tale of two banks that are very much out in front' — JPMorgan and Capital One.
Jul 29, 2025
NSF partnership on AI research institutes[11]
$1 million a year for five years supporting NSF's AI research ecosystem.
Jul 2, 2025
How the dealer agents were built[37]
At VB Transform, Naphade describes an evaluator agent trained on the bank's policies and a 55% lift in dealers' engagement and serious-lead metrics.
Jun 24, 2025
Patented AI, described[10]
The bank on its 'analytical DNA' and customer-centred AI patents.
Jun 4, 2025
ASKS Center for generative-AI safety at Illinois[33]
Capital One and the University of Illinois launch the Center for Generative AI Safety, Knowledge Systems and Cybersecurity; AI Foundations SVP Milind Naphade among its contributors.
May 18, 2025
Discover acquisition completes[2][21]
Adds more than 2,200 AI professionals to the talent pool, per Evident.
Mar 4, 2025
Chat Concierge launches for auto dealers[7][16]
A multi-agent assistant that compares vehicles, books appointments and hands off to sales, on the Auto Navigator platform.
Dec 1, 2024
Enterprise AI at scale, presented at AWS re:Invent[9]
The bank lays out use cases across AML, cybersecurity, digital and customer service, and its enterprise AI program.
Jun 3, 2021
Capital One Ventures invests in Socure[43]
Announcing the investment, Socure said Capital One had become 'one of our top clients'; Sarah Strauss, Capital One's head of fraud and servicing strategy, said Socure's technology had 'meaningfully increased our accuracy in identifying fraud', cutting customer friction while preventing additional fraud.
Sep 26, 2018
Why Eno's language technology was built in-house[4]
The bank says it built Eno's natural-language processing itself, from concept to launch in under three months, training it on customer-service chat logs; Eno understood more than 2,200 terms and emojis for a balance request.
Mar 10, 2017
Eno, a text-message chatbot, launches as a pilot[6][5]
Unveiled at SXSW as an SMS assistant that answers natural-language texts and emojis about balances and card payments; Capital One called it the first SMS text-based intelligent assistant from a US bank.

Where AI runs

Where does Capital One use AI today?

Figure 2 · Where AI runs at Capital One, and how far along. 13 of 13 named systems are in production; the rest are rolling out, piloted or announced. Status is as described in the cited source at its date.

System	What it does	Status	Use case
Chat Concierge[7]	Multi-agent assistant on dealer websites; compares vehicles, schedules appointments, routes to sales.	In production	Generative & agentic AI
AML and cybersecurity models[9]	Named by the bank among the areas where AI improves operations.	In production	AML / KYC
Eno[3][4]	Customer assistant in the mobile app, by text and email, in notifications (including on smart watches) and online from a computer: answers account questions and watches for charges that might surprise customers, free-trial endings and recurring charges. Natural-language processing built in-house.	In production	Customer-facing chatbots
Digital and customer-service AI[9]	Part of the enterprise AI program presented at re:Invent.	In production	Customer-facing chatbots
Brex agentic payments[19]	Agentic AI products for corporate cards and expenses, acquired with Brex.	In production	Generative & agentic AI
Databolt (Capital One Software)[42]	Tokenisation and data-security tooling sold to enterprises for secure AI adoption.	In production	Data & privacy
Fraudulent remote-worker detection[20]	The chief technology risk officer on nation-state impostor employees.	In production	Cybersecurity
MACAW fraud-call summaries[39]	Understanding, reasoning, validation and explanation agents on fine-tuned open-weight models summarise long fraud calls for several hundred specialist agents; the same architecture runs Chat Concierge.	In production	Fraud detection
Proprietary LLMs customised from open-source models[32][39]	Generative AI models built by customising open-source LLMs with synthetic and licensed open-source data; the disclosed training set exceeds 20 trillion tokens.	In production	Generative & agentic AI
ProtectID dealer fraud screen[36]	Proprietary fraud models and AI score every application sent through Dealer Navigator and flag synthetic and third-party fraud.	In production	Fraud detection
VulnHunter[38]	Agentic code-security tool with attacker-first analysis and a falsification engine; open-sourced in July 2026.	In production	Cybersecurity
AI and model risk in the 10-K[24]	The 2025 10-K names models and AI in pricing, fraud detection, loan grading, credit decisions and software development as a risk, plus agentic-commerce fraud and AI intellectual-property uncertainty.	In production	Model risk management
Socure identity-fraud screening (U.S. Card)[43][44]	Socure says its identity verification raised approval rates and cut false positives for the Capital One U.S. Card team (vendor-reported, 2021); Socure's company page still lists Capital One among organisations using it.	In production	Fraud detection

Use-case links open the tracker's rules-by-use-case matrix: which documents govern each system type.

AI vendors and fintech partners

Which AI vendors and fintechs does Capital One work with?

Company	What Capital One uses it for	Status	Category
Socure	Identity-fraud screening for the Capital One U.S. Card team, which Socure says raised approval rates and cut false positives; Capital One is also named on Socure's company page	In production	Fraud detection
Brex	Acquired Brex (closed April 2026); Brex's agentic AI for corporate cards and expenses	Acquired	Fintech · Business finance & infrastructure

Each vendor and fintech page cites the source for the relationship, and vendor pages list products, pricing and other bank customers. Compare AI tools for banks by category, or read the fintech and neobank AI dossiers.

The numbers

What has Capital One disclosed in numbers?

Metric	Value	As of
Evident AI index rank	#2 of 50 (three editions running)[18]	Oct 8, 2025
AI staff added	2,200+ (largely via Discover)[21]	Oct 9, 2025
Customers	100 million+[14]	Aug 3, 2026
Brex acquisition	$5.15 billion (announced); 35,000 clients[23][19]	Apr 7, 2026
University AI research commitments	$4.5 million (UVA); $1 million a year × 5 (NSF)[12][11]	Oct 27, 2025
Granted U.S. patents	5,000+[10]	Jun 24, 2025
AI patents granted in 2024	472 (up 14%); sixth overall and top bank on Harrity's 2025 AI Patent 100[40]	Dec 31, 2024
Dealer engagement and serious-lead lift reported with Chat Concierge	55%[37]	Jul 2, 2025
Generative-AI training data disclosed	20 trillion+ tokens; 7,100+ synthetic conversations[32]	Sep 24, 2026

In their words

“For companies, the threat of obsolescence is the greatest ever, and so too is the opportunity. The imperative to adapt and lead is existential.”[15]
Richard Fairbank · Founder and CEO · Mar 25, 2026

“Because at Capital One, we don't just use AI—we build AI.”[14]
Capital One · AI fluency program · Aug 3, 2026

“These opportunities require significant investment in AI and AI talent, and we are doing that.”[25]
Richard Fairbank · Founder, Chairman and CEO · Oct 21, 2025

“Our tech stack was built from the outset working backwards from the AI revolution and we are now building AI solutions across our businesses.”[26]
Richard Fairbank · Founder, Chairman and CEO · Jan 22, 2026

“All companies will be able to take advantage of AI, but the leverage is vastly greater when AI is embedded in the company's ecosystem.”[27]
Richard Fairbank · Founder, Chairman and CEO · Apr 21, 2026

“We felt an imperative to open-source VulnHunter because modern software supply chains are very connected, and the scale of the AI threat is larger than any single organization.”[38]
Chris Nims · EVP, Chief Information Security Officer · Jul 17, 2026

“Agentic AI now must be treated as an end-to-end system.”[45]
Rashmi Shetty · VP of Enterprise AI · Sep 28, 2026

“You're just massively reducing the stress in the system. It's more satisfying for everyone.”[46]
Prem Natarajan · EVP, Chief Scientist and Head of Enterprise AI · Sep 29, 2026

AI leadership

Who runs AI at Capital One?

Richard Fairbank

Founder, Chairman and CEO[15]

Prem NatarajanLinkedIn ↗

EVP, Chief Scientist and Head of Enterprise AI[1][21][22]

Andy OzmentLinkedIn ↗

EVP, Chief Technology Risk Officer[20]

Robert M. AlexanderLinkedIn ↗

Chief Information Officer[24]

Amy LenanderLinkedIn ↗

Chief Data Officer[35]

Milind NaphadeLinkedIn ↗

SVP, Technology, and Head of AI Foundations[37][33]

Chris Nims

EVP, Chief Information Security Officer[38]

Sanjiv Yajnik

President, Financial Services (Capital One Auto)[36][40]

Rashmi Shetty

VP of Enterprise AI[45]

Open AI roles

What AI roles is Capital One hiring for?

12 open AI-related postings on Capital One's careers site as of Oct 1, 2026. A snapshot of current openings, not a source: postings close.

Senior Director, Data Scientist - Global Payment Network

Riverwoods, IL (+3 more) · posted Sep 29, 2026

Director, AI Engineer (Remote Eligible)

San Francisco, CA (+6 more) · posted Sep 25, 2026

Director, Data Scientist - Generative AI Systems

McLean, VA (+1 more)

Director, AI Learning Enablement

McLean, VA (+1 more)

Business Director, Product Strategy & Analytics (Assistance & AI)

McLean, VA (+3 more) · posted Sep 2, 2026

Senior Staff AI Engineer - Agentic AI Platform (Remote Eligible)

San Francisco, CA (+5 more) · posted Sep 27, 2026

Senior Staff AI Engineer (Remote Eligible)

San Francisco, CA (+5 more) · posted Sep 23, 2026

Senior Staff Machine Learning Engineer

McLean, VA (+1 more) · posted Sep 29, 2026

Principal Data Scientist, AI Foundations

New York, NY (+2 more)

Principal Data Scientist - AI for Data and Assessment

McLean, VA (+2 more) · posted Sep 23, 2026

Principal Associate, Data Scientist - Applied AI

New York, NY (+1 more) · posted Sep 25, 2026

Principal Associate, Data Scientist - LLM Customization Team

New York, NY (+1 more)

Regulators

Which regulators govern Capital One's AI?

Figure 3 · The regulatory perimeter. Each authority links to its tracker page; the number is how many of its documents this page cites. Solid lines are the bank's direct supervisors; dashed lines are consumer, market and global standard-setters that reach it through its activities.

Authority	Why it matters here	Documents
OCC	National bank; the revised model risk guidance applies, and in-house models mean the bank owns validation end to end.	OCC Bulletin 2026-13
CFPB	A card and auto lender at scale: adverse-action duties, the 2026 Regulation B rule and the chatbot spotlight all bear directly on its AI.	ECOA / Regulation B adverse action (15 U.S.C. 1691(d); 12 CFR 1002.9) Regulation B final rule on disparate impact (April 2026) CFPB Chatbots in Consumer Finance (issue spotlight, 2023)
Colorado AI Act	Auto and card lending into Colorado: the ADMT Act's notice and explanation duties from January 1, 2027, and the conversational-AI disclosure law for tools like Chat Concierge.	SB 26-189 HB 26-1263
California CPPA	CCPA-scope business: the CPPA's automated decision-making regulations operative January 1, 2026.	CPPA ADMT, risk-assessment and cybersecurity-audit regulations
NIST	In-house model development maps to the NIST AI RMF and its generative-AI profile.	NIST AI RMF 1.0 NIST AI 600-1 (Generative AI Profile)

Side by side: SR 11-7 vs SR 26-2 · OCC vs CFPB on AI lending · PRA SS1/23 vs SR 26-2 · EU AI Act vs Colorado AI Act · NIST AI RMF vs ISO 42001 · all comparisons.

What the record suggests

What does the public record suggest about Capital One's AI strategy?

Analysis by BankingNewsAI from the sources cited on this page. Observations, not advice.

Building models means owning the model-risk burden

A bank that trains and customises its own LLMs cannot point to a vendor's validation. Under the revised guidance that is a heavier, not lighter, documentation load — and a reason the research posture doubles as a governance posture.

Chat Concierge is a consumer-law test case

An agent that steers a car buyer and schedules a sale sits close to state conversational-AI disclosure laws and CFPB expectations on chatbots. How it discloses itself and hands off to humans is the template others will copy.

The talent pool is the asset the acquisitions bought

Discover added 2,200 AI staff; Brex added agentic products. Integration risk and AI capability are now the same project.

The bank is becoming a vendor

Databolt and the software business put Capital One on the supply side of bank AI, with the third-party-risk questions reversed.

Latest AI news

Capital One’s agentic AI strategy hinges on data, platform-first mindset
ciodive.com · Sep 28, 2026
Capital One Financial (COF) Joins Banks Warning On AI Shopping Agent Risks
finance.yahoo.com · Sep 26, 2026
Why Capital One built its multi-agent AI platform around open-weight models
venturebeat.com · Aug 13, 2026
Advancing AI fluency at Capital One
capitalone.com · Aug 3, 2026

Capital One in the daily brief

Sep 23, 2026Six banks publish agentic commerce principles Jul 23, 2026Capital One has open-sourced an AI agent for finding code vulnerabilities Jul 19, 2026Capital One open-sourced an agentic vulnerability hunter Mar 24, 2026Capital One’s Databolt adds capabilities to tokenize unstructured data for AI use—signal that ‘AI-ready’ privacy engineering is productizing fast

Common questions

What is Eno, Capital One's chatbot?

Eno is Capital One's customer assistant. It was unveiled at SXSW in March 2017 as a text-message chatbot, which the bank described as the first SMS text-based intelligent assistant from a US bank, with natural-language processing built in-house and trained on customer-service chats. Today Eno is available in the Capital One mobile app, by text and email, through notifications (including on smart watches) and online from a computer: it answers account questions around the clock and alerts customers to charges that might surprise them, free-trial endings and recurring charges. Capital One has not published an interaction total for Eno.

What is Capital One's Chat Concierge?

Chat Concierge is Capital One's first proprietary multi-agent conversational AI assistant, launched in March 2025 for auto dealers through its Auto Navigator platform. Embedded on dealers' websites, it answers questions, compares vehicles, schedules test drives and passes qualified buyers to the sales team, taking actions rather than only providing information.

How does Capital One rank in AI among banks?

Evident's AI index, which scores the 50 largest banks in North America, Europe and Asia-Pacific on talent, innovation, leadership and transparency, has placed Capital One second behind JPMorgan Chase in three consecutive editions, most recently October 2025. Evident also puts its AI talent pool second, after the bank added more than 2,200 AI professionals through the Discover acquisition.

Who leads AI and technology at Capital One?

Founder, Chairman and CEO Richard Fairbank sets the AI strategy and discusses it on every earnings call. Prem Natarajan is EVP, Chief Scientist and Head of Enterprise AI; Milind Naphade is SVP and head of AI Foundations, the team behind its agentic systems; Robert M. Alexander is Chief Information Officer; Amy Lenander is Chief Data Officer; Chris Nims is Chief Information Security Officer; and Andy Ozment is Chief Technology Risk Officer. Sanjiv Yajnik, president of financial services, runs Capital One Auto, home of Chat Concierge and ProtectID.

Sources

Introducing Prem Natarajan
Capital One · May 24, 2023
Capital One Completes Acquisition of Discover
Capital One · May 18, 2025
Eno, your Capital One assistant
Capital One · Sep 19, 2026
How and Why Capital One Built Eno's NLP In-House
Capital One · Sep 26, 2018
Bot AI: How Capital One's Design Team Created SMS Bot Eno
Capital One · Sep 29, 2017
Capital One launches Eno, gender-neutral virtual assistant
Reuters · Mar 10, 2017
AI for Car Dealerships: Smarter Car Shopping
Capital One · Mar 5, 2025
Capital One's Applied AI research
Capital One · Nov 20, 2025
How AI is transforming financial services & banking
Capital One · Mar 27, 2025
Shaping Financial Services with Patented AI
Capital One · Jun 24, 2025
Capital One partners with NSF to advance U.S. AI leadership
Capital One · Jul 29, 2025
Capital One deepens UVA partnership to advance AI research
Capital One · Oct 27, 2025
Insights from the inaugural Capital One AI Symposium
Capital One · Apr 23, 2026
Advancing AI fluency at Capital One
Capital One · Aug 3, 2026
Annual report and shareholder letter (filed March 2026)
Capital One via SEC EDGAR · Mar 25, 2026
Capital One's 'Chat Concierge' Puts Agentic AI on Dealer Websites
The Financial Brand · Mar 4, 2025
Goldman Sachs, Capital One prep for self-driving AI agents
American Banker · Feb 20, 2025
JPMorganChase, Capital One top AI ranking, once again
American Banker · Oct 8, 2025
Capital One pushes deeper into AI, business payments (Brex closes)
American Banker · Apr 8, 2026
How North Korean operatives infiltrate US bank payrolls
American Banker · Mar 24, 2026
BofA, Capital One, Citi lead charge to scale AI talent pool
Banking Dive · Oct 9, 2025
Capital One on Building AI Moats in Banking (Tech Disruptors)
Bloomberg · Mar 17, 2026
Capital One Strikes $5.15 Billion Deal for Fintech Brex
The Wall Street Journal · Jan 22, 2026
Capital One Financial Corporation Annual Report on Form 10-K for 2025
Capital One via SEC EDGAR · Feb 19, 2026
Capital One (COF) Q3 2025 Earnings Call Transcript
The Motley Fool (Yahoo Finance) · Oct 21, 2025
Capital One (COF) Q4 2025 Earnings Call Transcript
The Motley Fool (Yahoo Finance) · Jan 23, 2026
Capital One (COF) Q1 2026 Earnings Transcript
The Motley Fool (Yahoo Finance) · Apr 22, 2026
Capital One Financial Corporation Q2 2026 Earnings Call Summary
Moby Intelligence (Yahoo Finance) · Jul 22, 2026
Capital One highlights Discover, Brex integration
American Banker · Jul 21, 2026
Capital One completes acquisition of Brex
Capital One · Apr 7, 2026
How Capital One is developing for the bank of the future
Capital One · Feb 7, 2022
GAI Training Data Transparency Disclosure
Capital One · Sep 24, 2026
Capital One and University of Illinois launch ASKS Center
Capital One · Jun 4, 2025
NLP research foundations at ICLR 2026
Capital One · Apr 20, 2026
Context is key to surpass AI's 40% success: Gartner D&A 2026
Capital One Software · Mar 26, 2026
Capital One Auto's ProtectID Wins 2026 'Fraud Prevention Innovation Award' for its Fraud Protection Tool
FinTech Breakthrough (GlobeNewswire) · Mar 19, 2026
Capital One builds agentic AI to supercharge auto sales
VentureBeat · Jul 2, 2025
Capital One releases VulnHunter, an open-source AI tool that finds software flaws before hackers do
VentureBeat · Jul 17, 2026
Why Capital One built its multi-agent AI platform around open-weight models (presented by Capital One)
VentureBeat (sponsored by Capital One) · Aug 13, 2026
Inside Capital One Auto's innovation, patent strategy
Bank Automation News · Aug 27, 2025
Are banks falling behind on agentic commerce standards?
American Banker · Sep 23, 2026
Capital One Software Expands Databolt Capabilities
Business Wire via FT Markets · Mar 23, 2026
Socure Increases Approval Rates and Slashes False Positives for Capital One U.S. Card Team as Capital One Ventures Completes Strategic Investment in Company
Socure · Jun 3, 2021
About Socure (undated page, retrieved 2026-09-24)
Socure · Sep 24, 2026
Capital One's agentic AI strategy hinges on data, platform-first mindset
CIO Dive · Sep 28, 2026
How bottom-line-focused financial firms are finding ROI in AI
Fortune · Sep 29, 2026

Other banks

Capital One:the AI strategy, sourced.