Webcast: Auditoria com IA: Tecnologias emergentes, desafios e futuro

Audit with AI: Emerging Technologies, Challenges, and the Future

Introduction to the Discussion

• The webcast focuses on the transformative role of AI in auditing, moderated by Raquel Brandim from Banco do Brasil.

• The discussion includes insights from experts Fábio, Eduardo, Paola, and Alexandre about the benefits of adopting AI and machine learning solutions within the bank.

Expert Introductions

• Raquel invites each expert to introduce themselves and their roles at Banco do Brasil.

• Fábio has 24 years at Banco do Brasil, with 17 years in internal auditing; he emphasizes innovation in audit practices.

• Eduardo has been with Banco do Brasil for 17 years as well; he works alongside Fábio focusing on analytics and generative AI.

• Paola has over six years in auditing after a decade in technology; she expresses enthusiasm for participating in the event.

• Alexandre joined Banco do Brasil in 2004 and moved into auditing in 2017; he currently works on methodology.

Historical Context of AI Implementation

• Raquel recalls an interview with Irã, the General Auditor of Banco do Brasil, discussing generative AI's implications for financial institutions.

• The conversation highlighted ethical considerations and implementation challenges regarding generative AI technologies within banks.

Accelerating Technology Adoption

• Fábio reflects on how Irã's challenge prompted them to accelerate their exploration of innovative technologies within internal auditing.

• He notes that Banco do Brasil has a long history of utilizing technology since the 1990s, positioning it as a market benchmark for data usage.

Cautious Approach to Generative AI

• Fábio discusses how they are carefully studying generative AI applications while maintaining a conservative approach typical of audit functions.

• He stresses that despite being cautious due to their nature as auditors, they must also embrace advanced tools to remain effective.

• The challenge lies in implementing generative technologies thoughtfully while managing associated risks transparently.

Challenges and Innovations in Auditing with Generative AI

Partnership and Initial Challenges

• The collaboration between the technology department and the auditing team was marked by a strong partnership, allowing for the implementation of initial use cases despite cultural and skill-related challenges.

• Transitioning to generative programming posed significant differences from traditional structured programming, requiring adaptation from the team.

Innovation in Auditing Practices

• The need for auditors to stay connected with advanced technologies is crucial for optimizing their actions within Brazil's sophisticated financial system.

• The audit team's innovative profile is reinforced by structured innovation programs that encourage creativity and adaptability in facing challenges.

Utilizing Generative AI for Risk Analysis

• Money laundering and terrorism financing are critical risks in financial auditing, necessitating thorough analysis as mandated by regulatory bodies.

• The integration of generative AI tools like GPT-4 has significantly enhanced data analysis capabilities, enabling auditors to process large volumes of information efficiently.

Efficiency Gains through Automation

• A new system utilizing GPT analyzes data related to money laundering risks, improving accuracy in identifying relevant issues during audits.

• This automation has reduced manual labor previously required for data analysis, leading to increased operational efficiency within the auditing process.

Human-AI Collaboration in Auditing

• The implementation of generative AI allows auditors to focus on more complex human-centric tasks while routine analyses are handled by machines.

• Generative AI reads every detail produced during audits with high accuracy, ensuring no critical information is overlooked.

Sentiment Analysis in Fraud Prevention

• The use of sentiment analysis via prompts enhances fraud detection processes; it streamlines operations compared to previous technologies that required extensive training.

Understanding AI in Auditing

The Role of AI in Contextual Understanding

• The use of databases allows AI to understand context easily, as it has a vast technical knowledge base. For example, when discussing money laundering, the AI can quickly assess relevant information without needing extensive training.

• The process is streamlined by asking the AI to evaluate topics for their positive or negative contributions to prevention and combatting practices, categorizing them accordingly.

Sentiment Analysis and Reporting

• The solution developed includes reading knowledge bases and performing sentiment analysis to categorize findings as positive, negative, or neutral regarding money laundering connections.

• A draft report is generated that highlights well-managed processes alongside areas needing improvement based on the sentiment analysis conducted by the AI.

Human Oversight in AI Classification

• As the AI classifies data, it provides justifications for its decisions. This human oversight ensures that there is continuous monitoring and adjustment of parameters to prevent inaccuracies.

• There’s an emphasis on understanding how generative AI differs from traditional uses of artificial intelligence, with ongoing learning about its impacts being crucial for effective implementation.

Current Applications of Artificial Intelligence

• Beyond generative solutions, other applications include combining big data processing with analytical platforms for comprehensive client monitoring projects (MBC).

• This project utilizes massive datasets from various systems—billions of records—to enhance auditing capabilities beyond mere sampling methods previously used due to limited data processing capacity.

Advancements in Machine Learning Technologies

• Automated machine learning (AutoML) technologies are being employed to improve efficiency in model selection and testing phases within projects.

• These advancements allow for automated treatment processes where models are chosen based on submitted variables, significantly reducing time spent on manual adjustments during machine learning projects.

Transformation Through Technology Integration

• The integration of new technologies across departments signifies a broader transformation within organizations. Projects like AutoML reflect long-desired capabilities finally becoming feasible due to improved technological infrastructure.

The Evolution of Predictive Modeling in Banking

Advancements in Classification and Predictive Modeling

• The emergence of advanced classification methods, like those used by GPT, has become feasible due to significant technological advancements that have reduced costs.

• Working with a large client base (90 million customers) and billions of transactions monthly presents immense challenges for compliance and operational efficiency.

Challenges in Financial Auditing

• The complexity of maintaining compliance with regulatory bodies is likened to a dream scenario for financial auditors, given the robust platforms available today.

• The integration of IT skills with auditing practices allows professionals to navigate both fields effectively.

Importance of Audit Planning

• Effective audit planning is crucial due to stringent regulatory requirements; tools like "planeja aí" are utilized for this purpose.

• Comprehensive mapping of processes, products, services, and risks is essential for auditors to identify high-impact risks efficiently.

Data Management and Strategy Alignment

• Close collaboration with the bank's strategy is necessary; auditors must gather external market intelligence alongside internal data.

• Understanding various market players (e.g., fintech companies, traditional banks) helps auditors contextualize their findings within the broader financial landscape.

Utilizing Generative AI in Auditing

• Generative AI can assist auditors by providing insights based on complex data sets from customer feedback and product performance.

• This technology enables auditors to link external information back into their auditing frameworks effectively.

Enhancing Auditor Decision-Making

• Generative AI acts as an assistant rather than replacing auditor judgment; it provides insights while leaving final decisions up to the auditor's expertise.

• The role of the auditor remains central; they leverage AI-generated insights but retain autonomy over their professional judgments.

Impact of Generative AI on Auditing Practices

Integration of Generative AI in Auditing

• The use of generative AI tools, like GPT, is seen as a significant ally for auditors, enhancing their capabilities and efficiency.

• Despite the integration of AI, human auditors remain essential for validating information and ensuring data accuracy, emphasizing the importance of human oversight in the auditing process.

Role of Human Oversight

• It is crucial to keep humans at the center of AI applications in auditing; auditors must validate outputs from AI systems to maintain quality.

• Since AI cannot be held accountable for decisions, it should serve as an assistant rather than a decision-maker within audit processes.

Training and Knowledge Sharing

• Continuous training and clear communication about the role of generative AI are vital to ensure that auditors understand its purpose as a supportive tool.

• The responsibility lies with auditors to make final decisions based on insights provided by AI, reinforcing that technology should not bear blame for outcomes.

Managing Expectations and Demand

• There is an increasing demand from audit teams for more advanced capabilities from generative AI tools; this requires careful management and realistic expectations regarding what can be delivered.

• The challenge lies in balancing technological advancements with existing operational capacities while ensuring that all team members are aligned with new implementations.

Community Maturity and Adaptation

• The auditing community shows maturity in adopting new technologies; there is a growing understanding among team members about how these tools can enhance their work.

• As familiarity with technology increases, so do suggestions for improvements and additional features from audit teams, indicating a proactive approach towards leveraging generative AI.

Future Perspectives on Technology Integration

• A cautious yet optimistic approach is taken towards integrating new technologies into auditing practices; learning curves are acknowledged as part of the process.

• Historical shifts towards digital collaboration (e.g., online meetings via Teams) illustrate how quickly adaptation can occur once technology becomes embedded in daily routines.

Exploring Predictive Capabilities and Data Analysis in Banking

The Role of Auditing in Technological Advancements

• Discussion on the predictive capabilities of data analysis, Big Data, and generative initiatives within banking.

• Emphasis on maintaining a strong relationship with technology departments (DTEC), highlighting the importance of collaboration in auditing processes.

• Acknowledgment of past challenges where auditing was seen as distanced from new technologies; however, this perspective is shifting towards more active participation.

Engaging with New Technologies

• Recent efforts to engage in pilot projects related to AI, indicating a proactive approach to integrating new technologies into auditing practices.

• Importance of staying close to digital acceleration movements within the bank to effectively adapt auditing strategies.

Future Directions for Auditing and Technology Integration

• Inquiry into future steps regarding AI and analytics integration within both auditing and broader banking operations.

• Recognition that methodologies are evolving to incorporate planning and operational assessments directly linked to technological advancements.

Enhancing Efficiency through Technology

• Discussion on how technology can streamline audit processes, allowing auditors to focus on strategic activities rather than operational tasks.

• Insight into how understanding processes is becoming more critical than coding skills, reflecting a shift towards product-oriented thinking.

The Impact of AI on Employment Dynamics

• Clarification that AI will not replace jobs but will transform work dynamics by reducing operational tasks while enhancing consultative roles.

• Highlighting the importance of human oversight in AI applications within banking, ensuring that human intellect remains central in decision-making processes.

Future of Auditing in Financial Institutions

Transformation and Independence in Auditing

• The auditing department is set to transform next year, focusing on external perspectives and new technologies while aligning with the bank's objectives to enhance internal processes and maximize results.

• Acknowledging the challenge of maintaining independence in auditing, which is crucial despite close partnerships with tech teams. This independence is essential for effective auditing practices.

• There’s a need to audit areas that were previously collaborative, emphasizing that all auditors must maintain objectivity even when assessing familiar departments or projects.

• The discussion highlights the importance of resolving the balance between collaboration and independence within the auditing process, indicating it as a significant yet manageable issue.

• Internal audit operates independently within the organizational structure, reporting directly to the Board of Directors, which ensures autonomy in auditing all institutional processes and systems.

Impact of Auditing on Business Quality

• The role of auditing is critical in regulated financial markets; effective audits lead to enhanced business quality by identifying potential issues before they escalate into problems.

• The work done by auditors not only anticipates problems but also improves efficiency and effectiveness in detecting fraud or other issues post-occurrence.

Invitation to Engage

• An invitation was extended for attendees to visit an upcoming digital event featuring various experiences, lectures, and panels related to finance and technology at a convention center.