>
¿Os imaginais un mundo en el que una IA haga un informe ejecutivo sobre el cumplimiento de una política o marco?
¿Hacer pruebas de auditoría hablando directamente con los datos?
¿Contar cada año con un risk assessment inicial fundamentado sobre datos históricos y alineado con los riesgos del Grupo?
>
>
>
<01. Rationale>
In a dynamic corporate environment, Internal Audit encounters challenges concerning the effective management of information, process automation, and compliance with evolving regulations. (IA)2 is an initiative designed to enhance Internal Audit by integrating artificial intelligence, automation, and predictive analytics to elevate quality, efficiency, and risk analysis. We have developed intelligent modules that perform a dual function: in the strategic realm, they conduct complex analyses by cross-referencing information, generating valuable insights, and proposing innovative pathways; while in the operational field, they significantly enhance auditors' productivity, optimizing processes and reducing repetitive tasks.
<02. Project Objective>
Develop an ecosystem of AI-based tools that enable internal auditors to optimize the review, analysis, and evaluation of risks in an automated and proactive manner, redefining Internal Audit in the next 2-3 years.
>
>
<03. Main components>
Agents
Regulatory Compliance
Experts in subjects and in management
Comprehensive Global AI Regulation Assessment
Automation
Routine IA Tasks
AI driven Audits
Reinforcing Continuous audit & monitoring
AI driven Risk Assessment
Facilitating the planning of audits
Resources
Infrastructure People, Training
>
>
<04. Implementation Plan>
Fase 2: 6 meses
Fase 4: 9 meses
Fase 1: 6 meses
Fase 3: 9 meses
Fase 5: 6 meses
AI R.Assesment, Training
Consolidation
Integration
AI Continuous Audit
Design & Plan
Execution of audits on 100% of transactions in real time. AI identifies drift patterns and anomalies, feeding insights back into strategic planning and enabling continuous monitoring.
First tools deployed. Integration with corporate systems via specialized pipelines to ensure broad data access across structured and unstructured sources.
Full automation and predictive auditing become operational. Continuous audits, training and certification define a new audit model, with major efficiency and quality gains.
Automation of risk evaluation using large volumes of diverse data. Progressively improves calibration, enabling dynamic audit planning. Includes auditor training on AI tools and methodologies.
Initial design and prototype development aligned with the AI Global Strategy. Modular and scalable architecture adaptable to multiple regions and regulations.
Strongly supported by the existing audiboxes ecosystem ecosystem
Alignment with AI Group Strategy
>
>
<05. E2E (IA)2 Audit>
We conducted a prototype of an audit with the strategic and operational modules already outlined in Brazil. The audit took approximately 4 hours to run, costing less than USD 15.
Previous Audit Reports, reommendations and auditor's paper
Knowledge Base
Internal Policies & External Regulations
Internal control information & customer complaints
Data lake data dictionary
In-depth risk analys
Audit planning & test matrix
code development & quality assessment
Run codes obtaining the results
>
>
Para que seamos más rápidos y contribuyamos cada vez más con la institución, ese es el camino...
Esta iniciativa conecta la auditoría con el futuro y nos posiciona a la vanguardia de la transformación...
No estamos creando una herramienta, estamos construyendo una nueva forma de auditar...
>
Thanks / Gracias
Modelos de innovación
AI driven Audits
AI driven Risk Assessment
Intelligent Audit Planning Automatic risk assessment based on historical and real-time data.
Risk scoring with automatic classification of priority audits.
Interactive dashboard with trend monitoring and early detection of anomalies. First approach of the four-year Audit Plan and the following year.
Automated Report Generation. AI collects data from multiple sources, analyzes it, and generates detailed reports with findings and recommendations. Natural Language Processing (NLP).AI converts unstructured data into actionable insights, structuring audit reports with relevant information. Integration with ERP and Compliance Systems.Autonomous auditing connects with internal systems to ensure adherence to internal and external regulations.
Continuous Monitoring and Anomaly Detection.Machine learning algorithms analyze transactions and behaviors in real-time to identify fraud and financial risks.
People & Training
Infrastructure
Technological Infrastructure & Software
Servers for on-premise or private cloud deployment.
Centralized and segmented databases for internal auditing.
AI development tools (TensorFlow, PyTorch, OpenAI API, SQL, Power BI). Access to ChatGPT Enterprise and Virtual Agent SaaS. Automation and data analysis tools (Python, R, etc.).
Beneficios:
✔ Mayor eficiencia en el uso de herramientas digitales.
✔ Reducción de la curva de aprendizaje en nuevas tecnologías.
✔ Capacitación continua en riesgos tecnológicos y auditoría basada en datos.
Objetivo:
Garantizar que los auditores internos adquieran las habilidades necesarias para trabajar con herramientas de IA, análisis de datos y tecnologías emergentes aplicadas a la auditoría.
Integration with Santander's Global Strategy
Relationship Model: Coordination with the Group's AI initiatives and with global teams and definition of common standards. Interoperability and Connectivity: Centralized platform, APIs and connectors and shared use of models. Synergy with ChatGPT Enterprise, SmartOCR and IDP and other iniciatives
Secure access, process automation, consistency and scalability. Shared Use Cases: Regulatory Chatbot, automation of summaries, predictive analysis, data extraction and classification.
Expert agents in day to day management
Example: How an audiBox works, How to close a recommendation in the system.
Subject Matter Experts (SMEs) Agents ....
Agents to help new auditors
Mentors guiding new auditors through their onboarding process
BuenosdIAs
IncidencIA
Connect to the mailbox and get a summary of the most relevant messages.
Ask using colloquial language what you want to analyze.
AnalIzA
Use generative IA for analyze dashboard and find out findings, recomendations and conclusions
Modelos de innovación
Comprehensive Global AI Regulation Assessment
AI regulations vary globally but generally permit controlled use. The EU leads with a risk-based framework (minimal to unacceptable risk classifications), setting standards others often follow. Most jurisdictions require human oversight, risk management, and ethical governance while allowing AI implementation with appropriate controls and supervision for regulatory compliance.
The EU AI Act risk assessment classifies applications into tiered categories, with this project likely falling under minimal to limited risk levels, requiring only basic governance measures that are readily achievable within the proposed structure.
Convención AI 2025
Juan Herrera Campos
Created on May 20, 2025
Start designing with a free template
Discover more than 1500 professional designs like these:
View
Discover Your AI Assistant
View
Urban Illustrated Presentation
View
Geographical Challenge: Drag to the map
View
Decisions and Behaviors in the Workplace
View
Tangram Game
View
Process Flow: Corporate Recruitment
View
Weekly Corporate Challenge
Explore all templates
Transcript
>
¿Os imaginais un mundo en el que una IA haga un informe ejecutivo sobre el cumplimiento de una política o marco?
¿Hacer pruebas de auditoría hablando directamente con los datos?
¿Contar cada año con un risk assessment inicial fundamentado sobre datos históricos y alineado con los riesgos del Grupo?
>
>
>
<01. Rationale>
In a dynamic corporate environment, Internal Audit encounters challenges concerning the effective management of information, process automation, and compliance with evolving regulations. (IA)2 is an initiative designed to enhance Internal Audit by integrating artificial intelligence, automation, and predictive analytics to elevate quality, efficiency, and risk analysis. We have developed intelligent modules that perform a dual function: in the strategic realm, they conduct complex analyses by cross-referencing information, generating valuable insights, and proposing innovative pathways; while in the operational field, they significantly enhance auditors' productivity, optimizing processes and reducing repetitive tasks.
<02. Project Objective>
Develop an ecosystem of AI-based tools that enable internal auditors to optimize the review, analysis, and evaluation of risks in an automated and proactive manner, redefining Internal Audit in the next 2-3 years.
>
>
<03. Main components>
Agents
Regulatory Compliance
Experts in subjects and in management
Comprehensive Global AI Regulation Assessment
Automation
Routine IA Tasks
AI driven Audits
Reinforcing Continuous audit & monitoring
AI driven Risk Assessment
Facilitating the planning of audits
Resources
Infrastructure People, Training
>
>
<04. Implementation Plan>
Fase 2: 6 meses
Fase 4: 9 meses
Fase 1: 6 meses
Fase 3: 9 meses
Fase 5: 6 meses
AI R.Assesment, Training
Consolidation
Integration
AI Continuous Audit
Design & Plan
Execution of audits on 100% of transactions in real time. AI identifies drift patterns and anomalies, feeding insights back into strategic planning and enabling continuous monitoring.
First tools deployed. Integration with corporate systems via specialized pipelines to ensure broad data access across structured and unstructured sources.
Full automation and predictive auditing become operational. Continuous audits, training and certification define a new audit model, with major efficiency and quality gains.
Automation of risk evaluation using large volumes of diverse data. Progressively improves calibration, enabling dynamic audit planning. Includes auditor training on AI tools and methodologies.
Initial design and prototype development aligned with the AI Global Strategy. Modular and scalable architecture adaptable to multiple regions and regulations.
Strongly supported by the existing audiboxes ecosystem ecosystem
Alignment with AI Group Strategy
>
>
<05. E2E (IA)2 Audit>
We conducted a prototype of an audit with the strategic and operational modules already outlined in Brazil. The audit took approximately 4 hours to run, costing less than USD 15.
Previous Audit Reports, reommendations and auditor's paper
Knowledge Base
Internal Policies & External Regulations
Internal control information & customer complaints
Data lake data dictionary
In-depth risk analys
Audit planning & test matrix
code development & quality assessment
Run codes obtaining the results
>
>
Para que seamos más rápidos y contribuyamos cada vez más con la institución, ese es el camino...
Esta iniciativa conecta la auditoría con el futuro y nos posiciona a la vanguardia de la transformación...
No estamos creando una herramienta, estamos construyendo una nueva forma de auditar...
>
Thanks / Gracias
Modelos de innovación
AI driven Audits
AI driven Risk Assessment
Intelligent Audit Planning Automatic risk assessment based on historical and real-time data. Risk scoring with automatic classification of priority audits. Interactive dashboard with trend monitoring and early detection of anomalies. First approach of the four-year Audit Plan and the following year.
Automated Report Generation. AI collects data from multiple sources, analyzes it, and generates detailed reports with findings and recommendations. Natural Language Processing (NLP).AI converts unstructured data into actionable insights, structuring audit reports with relevant information. Integration with ERP and Compliance Systems.Autonomous auditing connects with internal systems to ensure adherence to internal and external regulations. Continuous Monitoring and Anomaly Detection.Machine learning algorithms analyze transactions and behaviors in real-time to identify fraud and financial risks.
People & Training
Infrastructure
Technological Infrastructure & Software Servers for on-premise or private cloud deployment. Centralized and segmented databases for internal auditing. AI development tools (TensorFlow, PyTorch, OpenAI API, SQL, Power BI). Access to ChatGPT Enterprise and Virtual Agent SaaS. Automation and data analysis tools (Python, R, etc.).
Beneficios: ✔ Mayor eficiencia en el uso de herramientas digitales. ✔ Reducción de la curva de aprendizaje en nuevas tecnologías. ✔ Capacitación continua en riesgos tecnológicos y auditoría basada en datos.
Objetivo: Garantizar que los auditores internos adquieran las habilidades necesarias para trabajar con herramientas de IA, análisis de datos y tecnologías emergentes aplicadas a la auditoría.
Integration with Santander's Global Strategy
Relationship Model: Coordination with the Group's AI initiatives and with global teams and definition of common standards. Interoperability and Connectivity: Centralized platform, APIs and connectors and shared use of models. Synergy with ChatGPT Enterprise, SmartOCR and IDP and other iniciatives Secure access, process automation, consistency and scalability. Shared Use Cases: Regulatory Chatbot, automation of summaries, predictive analysis, data extraction and classification.
Expert agents in day to day management
Example: How an audiBox works, How to close a recommendation in the system.
Subject Matter Experts (SMEs) Agents ....
Agents to help new auditors
Mentors guiding new auditors through their onboarding process
BuenosdIAs
IncidencIA
Connect to the mailbox and get a summary of the most relevant messages.
Ask using colloquial language what you want to analyze.
AnalIzA
Use generative IA for analyze dashboard and find out findings, recomendations and conclusions
Modelos de innovación
Comprehensive Global AI Regulation Assessment
AI regulations vary globally but generally permit controlled use. The EU leads with a risk-based framework (minimal to unacceptable risk classifications), setting standards others often follow. Most jurisdictions require human oversight, risk management, and ethical governance while allowing AI implementation with appropriate controls and supervision for regulatory compliance. The EU AI Act risk assessment classifies applications into tiered categories, with this project likely falling under minimal to limited risk levels, requiring only basic governance measures that are readily achievable within the proposed structure.