In early 2023, Google LLC (Google) faced a pivotal decision: how to scale its use of artificial intelligence (AI) and machine learning (ML) models in financial forecasting without compromising regulatory compliance and internal control integrity. While AI/ML tools (such as predictive analytics, document parsing, and chatbot-based automation) improved efficiency, they also introduced risks such as data completeness and accuracy issues, model drift due to continuous learning, lack of transparency in decision-making, and challenges in reproducing results for audit purposes. Google had to design a risk and control strategy that balanced innovation and compliance with the Sarbanes-Oxley Act of 2002 (SOX). How could Google integrate AI/ML models into its SOX-regulated financial systems? How could it use cross-functional collaboration to overcome implementation challenges and drive both innovation and effective governance? And how could it meet future regulatory standards while reinforcing investor trust?
治理创新：谷歌在金融系统中对人工智能/机器学习的《萨班斯法案》控制措施 (Governing Innovation: Google’s SOX Controls for AI/ML in Financial Systems Simplified Chinese version)
Glorin Sebastian, Eshan Bhatt
Product #:W49924
Supplier:Ivey
Discipline:Accounting, Finance
Setting:United States, 2024
Your Price:$10.54
Purchase this material with a free Educator Account and enjoy more benefits:
- Free Educator Copy
- Free Teaching Notes
- Build your coursepack
- Discounted academic price
Learning Objectives
- 理解人工智能/机器学习模型在财务报告中引入的独特风险。
- 分析传统控制框架如何演进以应对概率模型。
- 评估跨职能协作如何支持可扩展治理。
- 识别人工智能/机器学习模型生命周期控制的最佳实践。
- 设计符合监管要求的基于风险的控制策略。
This case can be used in the following courses: finance or accounting information systems, or corporate governance (upper-year bachelor of business administration), internal controls and compliance (graduate level), information systems audit (graduate and executive MBA level), and corporate governance for technology-driven firms (graduate or PhD level). After working through the case and assignment questions, students will be able to do the following:
- Understand the unique risks introduced by AI/ML models in financial reporting.
- Analyze how traditional control frameworks must evolve to address probabilistic models.
- Evaluate how cross-functional collaboration can support scalable governance.
- Identify best practices for AI/ML model lifecycle controls. Design risk-based control strategies aligned with regulatory requirements.