Banorte, a Mexican banking, insurance and pension group, runs AI as an operating-model programme led by its chief operating officer: years of machine-learning personalisation, a Maya virtual assistant handling more than 2 million messages a month, and a 2026 plan to put AI into the daily work of 10,000 operations and back-office staff and build toward one AI agent per client.
How is Banorte using AI?
Grupo Financiero Banorte (GFNorte), the Mexican group with Ps 2,793,578 million in total assets at 30 September 2026, treats AI as a way to change how the bank operates rather than as a product line. Its 2025 annual report describes machine-learning personalisation, a Maya virtual assistant that fields more than 2 million messages a month, specialised internal chatbots, a framework for retrieval-augmented generation and a Microsoft Copilot rollout after a 350-person pilot. In April 2026 CEO Marcos Ramírez said the bank intends to have 10,000 people in operations and back-office areas integrating daily work with AI, under a formal AI governance structure, and COO Rafael Arana described a plan for one AI agent per client. Since July 2026 COO Rafael Arana has led the group's transformation, technology and the implementation of AI initiatives. Management links AI to its cost-to-income goal of about 34%. Partners on record include Google Cloud (since 2021), Red Hat OpenShift AI, Hitachi Vantara and SAP Fioneer. In the third quarter of 2026 Banorte said it was implementing an internal control framework for AI that addresses hallucinations, bias, data exposure, third-party dependence and agent autonomy.
Lead bank
Banco Mercantil del Norte, S.A. (Banorte)
Total assets (group, as reported, Sep 30, 2026)
Ps 2,793,578 million total assets (GFNorte consolidated)[2]
Charter
Mexican financial group (Grupo Financiero Banorte, S.A.B. de C.V.) whose main subsidiary, Banco Mercantil del Norte, is a multiple banking institution supervised by the Comisión Nacional Bancaria y de Valores (CNBV) and Banco de México; the group also owns insurance, annuities, brokerage and the Afore XXI Banorte pension fund
Headquarters
San Pedro Garza García, Nuevo León, Mexico
Flagship platform
Hyper-personalisation analytics, Maya, Red Hat OpenShift AI and Google Cloud
AI leadership
José Marcos Ramírez Miguel · Rafael Arana de la Garza · Tomás Lozano Derbez · Olga Ceja · Delfin Ruiz · Jesús Ricardo Medina Cantú · Gerardo Salazar Viezca · José Francisco Martha González · Carlos Hank González
- Scale of staff AI: the 1Q26 plan is for 10,000 people across operations and back office to integrate daily activities with AI at 'levels 1 and 2', with some agents for more complex tasks, guided by a formal AI governance structure (CEO and COO, 22 April 2026).
- Customer AI: Maya, the virtual assistant on the web portal and Banorte Móvil, fields more than 2 million messages a month; machine-learning models drive payment reminders, recommendations based on consumption habits and emotion detection in assisted channels (2025 annual report).
- Agent per client: on the 1Q26 call (22 April 2026) COO Rafael Arana described a plan for an agent per client that manages many other agents; in November 2026 CEO Marcos Ramírez said the bank keeps moving toward its vision of one agent per client and that its approach combines autonomous AI capabilities with human judgment.
- Efficiency link: COO Rafael Arana tied the deployment of bank-wide AI tools to bringing the cost-to-income ratio back toward 34% (4Q25 call); the 2025 efficiency ratio was 35.8%.
- Measured results: Elina cut average contact-centre call time from 10 minutes to 50 seconds and agent training from 37 to 14 hours (The Banker, May 2025); Banorte says its automation and AI initiatives contributed to a tenfold rise in product acceptance, a 70% cut in attrition and more than seven points of NPS (3Q26 report).
- Generative AI stack: Red Hat OpenShift AI underpins Banorte's AI and GenAI architecture, with RAG assistants for incident diagnosis and IT requirements (Red Hat, April 2026); Microsoft Copilot was authorised for widespread use after a 350-person pilot (2025 annual report).
- Governance: the Code of Conduct gained guidelines on responsible use of AI tools in 2025, the board received AI and cybersecurity training, and an AI internal control framework was being implemented in 3Q26.
- Leaders: CEO Marcos Ramírez Miguel; COO Rafael Arana de la Garza (transformation, technology and AI since July 2026); Olga Ceja, Executive Director of Artificial Intelligence and Innovation; Delfin Ruiz, chief technology infrastructure officer.
- Digital-bank strategy changed: Banorte launched the fully digital bineo in January 2024 on SAP Fioneer's cloud platform, agreed to sell it to Klar's Clearscope in September 2025, and completed the purchase of Rappi's stake in RappiCard issuer Tarjetas del Futuro in December 2025.
What is Hyper-personalisation analytics, Maya, Red Hat OpenShift AI and Google Cloud?
Banorte's customer-facing AI is its hyper-personalisation engine (machine-learning models that drive offers, reminders and emotion detection in assisted channels) and Maya, the virtual assistant on its web portal and Banorte Móvil, which fields more than 2 million messages a month. For staff it built Elina, an AI contact-centre tool that consolidates more than 25 systems, specialised internal chatbots, a framework for retrieval-augmented generation (RAG) solutions with synthetic accuracy tests, and a Microsoft Copilot rollout after a 350-person pilot. Its generative AI architecture runs on Red Hat OpenShift AI, which Banorte adopted after five years on OpenShift; Google Cloud has been its strategic cloud partner since June 2021, an alliance renewed and expanded in 2025 for AI, advanced analytics and cloud cybersecurity. An internal control framework for AI adoption was being implemented in the third quarter of 2026.[1][17][19][8][2]
Which technology partners does Banorte work with on AI and cloud?
Banorte's public record names several partners rather than one AI provider. Google Cloud has been a strategic partner since June 2021 and the alliance was expanded in 2025 to cover AI, advanced analytics and cloud cybersecurity. Red Hat OpenShift AI hosts its AI and generative AI architecture, Microsoft Copilot was authorised for staff after a 350-person pilot, and Hitachi Vantara storage supports its relocated mainframe. The sources reviewed do not name the language-model providers behind Maya or the RAG assistants.[8][1][19][18]
| Partner | What it is | Status | Date |
|---|---|---|---|
| Google Cloud | Strategic cloud alliance; DORA programme for software delivery; expanded in 2025 for AI, analytics and cloud cybersecurity | In production | 2021-06-02[8][15][1] |
| Red Hat | OpenShift AI as the base of the AI and GenAI architecture, with RAG assistants | In production | 2026-04-23[19] |
| Microsoft | Copilot365 pilot for 350 people, then Copilot Chat authorised for widespread use | Rolling out | 2025[1] |
| Hitachi Vantara | Mainframe storage and data-centre migration to Querétaro | In production | 2026-03-04[18] |
| SAP Fioneer | Cloud banking platform for bineo (agreed for sale in 2025) | In production | 2024-02-02[21][12] |
| Volante Technologies | VolPay cross-border payments platform | Announced | 2020-10-21[22] |
What has Banorte done on AI, and when?
- Nov 4, 2026
Q3 2026: AI in coding and human judgment[4]
CEO Marcos Ramírez said Banorte is using AI across core business segments to simplify workflows and automate tasks, that coding tools are already improving development speed, and that its approach combines autonomous AI with human judgment.
- Nov 3, 2026
AI internal control framework under way[2]
The 3Q26 report says an internal control framework for safe, responsible adoption of AI is being implemented, in response to risks such as hallucinations, bias, data exposure, third-party dependencies and agent autonomy.
- Sep 1, 2026
Banorte cites AI results on inclusion in Evident's LatAm index[2]
Banorte's 3Q26 report says it was the only Mexican bank in the Evident AI Index for Banks LATAM, and that its automation and AI initiatives contributed to a tenfold increase in product acceptance, a 70% reduction in attrition and more than seven points of NPS.
- Jul 21, 2026
COO given responsibility for AI implementation[7]
Tomás Lozano became CFO reporting to COO Rafael Arana, who continues to lead the group's transformation, finance, strategy, technology and the implementation of AI initiatives.
- Apr 23, 2026
Red Hat recognises OpenShift AI deployment[19]
Red Hat gave Banorte a Latin America honourable mention for adopting Red Hat OpenShift AI as the basis of its AI and GenAI architecture, with RAG assistants that diagnose production incidents and draft functional requirements for IT projects.
- Apr 22, 2026
Plan for 10,000 staff working with AI and one agent per client[5]
CEO Marcos Ramírez said Banorte intends to have 10,000 people in operations and back office integrating daily work with AI under a formal governance structure; Rafael Arana said back-office work should run on AI by year-end and described an agent per client that manages other agents.
- Mar 4, 2026
Mainframe data centre moved with Hitachi Vantara[18]
Banorte relocated its mainframe data centre from Mexico City to Querétaro on Hitachi Vantara storage, migrating 450 terabytes in under an hour; transaction response times fell 50% and nightly batch processing shortened by an hour.
- Jan 28, 2026
AI spending accelerates, tied to the efficiency target[6]
On the 4Q25 call Rafael Arana said investment in AI had accelerated and teams had to be reinforced, and that 2026 would see the deployment of bank-wide AI tools as part of returning the cost-to-income ratio toward 34%.
- Dec 17, 2025
- Sep 4, 2025
Agreement to sell bineo to Klar's Clearscope[12]
GFNorte agreed to sell all shares of Banco Bineo to Clearscope Holdings, a subsidiary of Klar, subject to approvals from Mexico's finance ministry, CNBV, Banco de México and the antitrust authority.
- May 20, 2025
- Apr 15, 2025
Agreement to buy Rappi's stake in RappiCard issuer TDF[10]
GFNorte agreed to buy the 44.28% of Tarjetas del Futuro that Rappi would hold, for USD 50 million, with an exclusive 15-year commercial agreement with Rappi, to grow cross-selling powered by hyper-personalisation; TDF had more than 1.14 million cardholders at end-2024.
- Dec 17, 2024
Google Cloud DORA award for software delivery[15]
GFNorte won Google Cloud's 2024 DORA award for large-scale improvements: daily software releases instead of bi-weekly cycles, change lead time under two days, 20% lower development and operations costs and a projected 30% cut in infrastructure costs.
- Feb 2, 2024
- Jan 29, 2024
- Jun 2, 2021
Strategic alliance with Google Cloud[8]
Banorte announced a strategic alliance with Google Cloud to work together on the bank's digital transformation and the personalised development of new products and services.
- Oct 21, 2020
Cross-border payments modernisation with Volante[22]
Volante Technologies said Banorte would use its VolPay platform to modernise its corporate and retail cross-border payments processing.
- Jun 23, 2020
Digital financial services venture with Rappi[9]
GFNorte signed an alliance with Rappi for a 50-50 company offering digital financial services to Rappi's customers, committing up to 4 billion pesos over 18 months, as part of its digital and technological strategy.
Where does Banorte use AI today?
| System | What it does | Status | Use case |
|---|---|---|---|
| Maya virtual assistant[1] | Virtual assistant on the web portal and Banorte Móvil that fields more than 2 million messages a month; its NPS rose two points in 2025, and a 'Maya Virtual Supervisor' project won a regional contact-centre award. | In production | Customer-facing chatbots |
| Elina AI contact-centre tool[17][16] | Unified AI tool for contact-centre agents that consolidates more than 25 systems; average call time fell from 10 minutes to 50 seconds and agent training from 37 to 14 hours. | In production | Generative & agentic AI |
| Microsoft Copilot for employees[1] | A 350-person Copilot365 pilot produced time savings on day-to-day tasks; Copilot Chat was then authorised for widespread use across the organisation. | Rolling out | Generative & agentic AI |
| AI in the daily work of 10,000 operations and back-office staff[5][6] | A plan, not a reported result: on 22 April 2026 CEO Marcos Ramírez said the intention is to have 10,000 people across operations and back office integrating daily activities with AI 'levels 1 and 2', under a formal AI governance structure; COO Rafael Arana said back-office staff should be managing their work with AI ('level 1') by the end of the year, with some agents ('level 2') for more complex tasks. | Announced | Generative & agentic AI |
| One AI agent per client[5][4] | Target design in which a client-level agent orchestrates other agents, built on real-time data and lifetime-value models; described as a vision the bank keeps moving toward. | Announced | Generative & agentic AI |
| RAG assistants on Red Hat OpenShift AI[19][1] | Retrieval-augmented assistants over internal knowledge bases that diagnose production incidents and generate functional requirements for IT projects; a RAG framework uses synthetic tests for accuracy and consistency plus security testing. | In production | Generative & agentic AI |
| AI coding tools[4] | Management says coding tools are already improving development speed and creating opportunities for cost savings. | In production | AI-generated code & coding agents |
| AI and machine learning in credit risk[1] | AI and machine learning were incorporated into credit risk tools to raise model accuracy, alongside a new retail decision engine that uses alternative data and strengthens origination fraud detection. | In production | Credit scoring & underwriting |
| Hyper-personalisation and emotion models[1][2] | Machine-learning models send payment reminders and recommendations based on consumption habits, and models translate customer feedback into emotions in assisted channels to support cross-selling. | In production | Customer-facing chatbots |
| AI climate-risk monitor[1] | An AI-based monitor evaluates acute and chronic physical climate risks across the wholesale credit portfolio. | In production | AI governance (general) |
| Machine learning in threat detection[1] | Machine-learning solutions added to database monitoring and protection, with process automations for threat detection and containment. | In production | Cybersecurity |
| AI internal control framework[2] | Framework for safe, responsible and structured AI adoption covering hallucinations, bias, data exposure, third-party dependencies and agent autonomy. | Rolling out | Model risk management |
| Google Cloud alliance[8][1][15] | Strategic cloud partner since June 2021; the alliance was renewed and expanded in 2025 to cover AI, advanced analytics and cloud cybersecurity; the DORA programme brought daily releases and 20% lower development and operations costs. | In production | Third-party & vendor AI |
Use-case links open the tracker's rules-by-use-case matrix: which documents govern each system type.
What has Banorte disclosed in numbers?
| Metric | Value | As of |
|---|---|---|
| Total assets (GFNorte) | Ps 2,793,578 million[2] | Sep 30, 2026 |
| Staff to integrate daily work with AI | 10,000 in operations and back office (target)[5] | Apr 22, 2026 |
| Maya messages handled | More than 2 million a month[1] | Dec 31, 2025 |
| Copilot pilot | 350 employees[1] | Dec 31, 2025 |
| Elina average call time | From 10 minutes to 50 seconds[17] | May 19, 2025 |
| Elina agent training time | From 37 to 14 hours[17] | May 19, 2025 |
| Effects Banorte attributes to automation and AI | Tenfold rise in product acceptance, 70% lower attrition, NPS up more than seven points[2] | Sep 30, 2026 |
| Market-research time at Operadora de Fondos with AI tools | Down 83%, from 30 minutes to five a day[1] | Dec 31, 2025 |
| Efficiency ratio | 35.8% in 2025; management target about 34%[1][6] | Dec 31, 2025 |
| Development and operations cost reduction (Google Cloud DORA) | 20%; infrastructure costs projected 30% lower[15] | Dec 17, 2024 |
| Mainframe data migrated (Hitachi Vantara) | 450 terabytes in under an hour; response times down 50%[18] | Mar 4, 2026 |
| Banking clients | 13.8 million (8.1 million Afore clients)[3] | Sep 30, 2026 |
“We were among the first institutions to adopt AI at scale and our intention is to have 10,000 people across operations and back office areas, integrating daily activities with artificial intelligence levels 1 and 2.”[5]
Marcos Ramírez Miguel · Chief Executive Officer · Apr 22, 2026
“Importantly, our approach combines autonomous AI capabilities with human judgment.”[4]
Marcos Ramírez Miguel · Chief Executive Officer · Nov 4, 2026
“We don't need to buy more technology, but we have to use more of the technology that we have”[6]
Rafael Arana de la Garza · Chief Financial and Operating Officer · Jan 28, 2026
“Our digital capabilities are crucial to drive this strategy, as they will allow us to deliver hyper-personalized customer experiences,”[18]
Delfin Ruiz · Chief Technology Infrastructure Officer · Mar 4, 2026
“We use artificial intelligence to offer innovative and customizable products and functionalities that adapt to their needs and lifestyle.”[14]
Francisco Martha · CEO of Digital Business Development · Feb 1, 2024
“Two years ago we joined the Google Cloud DORA program to strengthen our ability to innovate, accelerate changes and transform platforms and applications throughout Grupo Financiero Banorte.”[15]
Carlos Hank González · Chairman of the Board · Dec 17, 2024
“There's a segment of our clients that today demands a 100% digital service,”[20]
Carlos Hank González · Chairman of the Board · Jan 29, 2024
Who runs AI at Banorte?
José Marcos Ramírez Miguel
Chief Executive Officer, Grupo Financiero Banorte; presents the AI programme to investors[1][5][4]
Rafael Arana de la Garza
Chief Operating Officer; leads the group's transformation, finance, strategy, technology and the implementation of AI initiatives (as of 21 July 2026; previously Chief Financial and Operating Officer)[7][6]
Tomás Lozano Derbez
Chief Financial Officer from 21 July 2026, reporting to the COO; previously head of Investor Relations, Corporate Development, ESG and Financial Planning[7]
Olga Ceja
Executive Director of Artificial Intelligence and Innovation; received the Fintech Americas 'AI and Automation' Platinum Award for developing Maya (2025)[1]
Delfin Ruiz
Chief Technology Infrastructure Officer (led the mainframe data-centre migration with Hitachi Vantara)[18]
Jesús Ricardo Medina Cantú
Chief Information Security Officer since 2018 (per an undated Banorte profile), reporting to the CEO and to the Audit Committee[13]
Gerardo Salazar Viezca
Managing Director, Risk Management and Credit (risk models with AI and machine learning)[1][4]
Which regulators govern Banorte's AI?
| Authority | Why it matters here | Documents |
|---|---|---|
| Comisión Nacional Bancaria y de Valores (CNBV) and Banco de México | Banorte's home supervisors. Mexico has no AI-specific banking rule on the record reviewed; the CNBV certifies Banorte's internal reserve and capital models, Mexican rules since 2018 require a CISO reporting to the CEO, and the CNBV and Banco de México had to approve the bineo sale. Banorte's 3Q26 report says its AI internal control framework is meant to keep AI use within regulatory compliance. | |
| Basel Committee | Dependence on Google Cloud, Red Hat, Microsoft and Hitachi Vantara for AI and core infrastructure engages the Basel Committee's third-party risk principles, and real-time client data for the agent-per-client design engages BCBS 239 risk-data principles. | BCBS Third-Party Risk Principles (Dec 2025) BCBS 239 |
What does the public record suggest about Banorte's AI strategy?
Analysis by BankingNewsAI from the sources cited on this page. Observations, not advice.
01
AI is framed as an efficiency programme
Management ties AI directly to bringing the cost-to-income ratio back toward 34% from 35.8% in 2025, and calls 2025 and 2026 'in between years'. The 10,000-person AI rollout in operations and back office is the measure to track against that ratio in 2027.
02
Outcome figures exist, but few are per use case
Banorte publishes some hard results (Elina's 50-second calls, an 83% cut in market-research time, a tenfold rise in product acceptance it attributes partly to AI) but no AI value figure or model count. The agent-per-client design is still described as a vision.
03
A mixed vendor stack rather than one AI partner
Google Cloud is the long-standing cloud partner, Red Hat OpenShift AI hosts the generative AI architecture, Microsoft supplies Copilot and the mainframe stays on premises with Hitachi Vantara storage. Which model providers sit behind Maya and the RAG assistants is not disclosed.
04
Digital strategy moved from a separate bank to the core
Banorte launched bineo in 2024 as a bank without legacy systems, then agreed to sell it in 2025 and bought Rappi's stake in RappiCard's issuer instead. AI personalisation inside the main bank, rather than a stand-alone digital bank, is now the centre of the strategy.
Which leader rankings does Banorte appear in?
6 of the computed leaders rankings include Banorte, each ranked from the same sourced records (bank pages, filings, hiring snapshots, the agent tracker, the incidents register).
How does Banorte compare with its peers?
- How does Grupo Financiero Banorte compare with its peers on AI and technology?
- Grupo Financiero Banorte competitors and alternatives, by business line
Its peer group on this site, by business model and size:
How is Banorte using AI?
Banorte uses machine learning for hyper-personalised offers, reminders and emotion detection; the Maya virtual assistant, which fields more than 2 million messages a month; the Elina contact-centre tool; internal chatbots and RAG assistants on Red Hat OpenShift AI; Microsoft Copilot for staff; and AI in credit risk, climate-risk monitoring and cybersecurity. In April 2026 its CEO said the intention is to have 10,000 operations and back-office staff integrating daily work with AI.
Who leads AI at Banorte?
Chief Operating Officer Rafael Arana de la Garza has led the group's transformation, technology and implementation of AI initiatives since a July 2026 reorganisation. CEO Marcos Ramírez Miguel presents the AI plan to investors, and Olga Ceja is Executive Director of Artificial Intelligence and Innovation, credited with developing Maya.
What is Banorte's 'one agent per client' plan?
Banorte's management describes a goal of giving each client an AI agent that coordinates other agents to serve that client's product mix, built on real-time data and lifetime-value models. In November 2026 the CEO called it a vision the bank keeps moving toward, combined with human judgment.
Is Banorte cutting jobs because of AI?
Banorte has not announced AI-related job cuts. Management talks about reinforcing AI teams, about 10,000 staff working with AI, and about a human-digital operating model in which people remain at the centre of decisions, while linking AI to a lower cost-to-income ratio.
Which cloud and AI vendors does Banorte use?
Google Cloud (strategic alliance since 2021, expanded in 2025 for AI and analytics), Red Hat OpenShift AI for its generative AI architecture, Microsoft Copilot for employees, Hitachi Vantara storage for its mainframe, and SAP Fioneer's cloud banking platform for bineo.
What is Maya?
Maya is Banorte's virtual assistant on its web portal and the Banorte Móvil app. The 2025 annual report says it fields more than 2 million messages a month, and its Net Promoter Score rose two points in 2025.
How does Banorte govern AI?
Banorte says a formal AI governance structure guides its rollout. In 2025 it added guidelines on responsible AI use to its Code of Conduct and trained its board on AI and cybersecurity, and in 2026 it was implementing an internal control framework that addresses hallucinations, bias, data exposure, third-party dependencies and agent autonomy.
- Integrated Annual Report 2025 (undated: no publication date on the report or the IR page; the 1Q26 call says it was published in the first quarter of 2026; retrieved 2026-11-19)
- 3Q26 Financial Results
- Leading Bank in Mexico 3Q26: Executive Summary
- Earnings call transcript: Grupo Financiero Banorte posts mixed Q3 2026 results
- Q1 2026 Grupo Financiero Banorte Earnings Call transcript
- Q4 2025 Grupo Financiero Banorte Earnings Call transcript
- Material Event: Changes in Organizational Structure
- Press Release 2Q21: Banorte reports 7% growth in consumer loans (Recent events: alliance with Google Cloud)
- Material Event: GFNorte agrees to strategic alliance with Rappi
- Material Event: acquisition of Tarjetas del Futuro and commercial agreement with Rappi
- Material Event: Acquisition of Tarjetas del Futuro and Execution of a Strategic Commercial Agreement
- Material Event: GFNorte entered into an agreement to sell all the shares representing the capital stock of Bineo
- Chief Information Security Officer (CISO) Profile 2024 (undated; posted with the 2025 sustainability documents; retrieved 2026-11-19)
- Banorte Chairman, Carlos Hank Gonzalez, launches bineo, the first 100% digital bank in Mexico
- Banorte, chaired by Carlos Hank González, is recognized by Google Cloud for accelerating digital innovation in Mexican banking
- Banorte, led by Chairman Carlos Hank-González, named best Latin American bank for innovation and technology by The Banker
- Technology Awards 2025 - Latin America - Banorte
- Banorte Partners with Hitachi Vantara to Build a Future-Ready Digital Banking Platform
- Red Hat Announces 2026 Honorable Mentions for Open Source Projects in Latin America
- Mexico's Banorte launches digital bank bineo
- Mexican banking heavyweight Banorte taps SAP Fioneer to launch digital bank Bineo
- Banorte Partners with Volante Technologies For Payments Modernization
When Banorte moves on AI, you'll read it here first.
when one of the banks we cover moves on AI, it is in the next morning's brief · six sourced stories · 7 am PT · free
plus every regulation, bank and agent page update, the morning after · leave any morning