Is Analytics Dead? The Truth About AI’s Role in Modern Audit

I was having lunch with a Chief Audit Executive (CAE) recently, and an interesting question came up: “Is analytics dead (i.e., should we stop doing analytics)? Is it all about AI now?” 

The question highlights a growing confusion around the role of analytics in an age dominated by artificial intelligence. The truth is, analytics certainly isn’t dead — but it is evolving. While some audit leaders may think they need to choose between AI and analytics or that AI will require hiring data scientists to be effective, the reality is that the two fields are complementary and should be used by all auditors. In this article, I debunk common myths surrounding analytics and AI while demonstrating how AI can supercharge your audit analytics program.

Myth 1: AI Replaces Analytics

The misconception that AI makes analytics obsolete is widespread. However, like automation supports analytics, AI complements rather than replaces it. Automation helps streamline repetitive tasks, while analytics provides critical insights into organizational data. AI enhances these insights by making processes faster and more efficient. Analytics remains unparalleled in uncovering trends and outliers. For example, by running outlier analyses, auditors can discover anomalies that traditional methods might miss. These insights often lead to questions organizations didn’t know they needed to ask. While it is powerful, AI can’t yet replace the nuanced investigative work of an auditor equipped with data analytics, but it can amplify its impact.

Myth 2: AI Is Limited to Text-Based Applications

A common assumption is that AI’s usefulness in an internal audit context is restricted to text-heavy tasks like report generation, risk control matrices (RCMs), and audit planning. While these applications are transformative, AI’s potential goes far beyond. Emerging AI tools can analyze public datasets, identify patterns, and generate interactive dashboards. Imagine feeding a dataset into an AI model and receiving real-time actionable insights. Although these advanced capabilities are not yet widespread (and may trigger fears about data privacy — see myth 4 below!), they are on the horizon. Organizations should be prepared to leverage these tools as they mature.

Myth 3: You Need Coding Skills to Use AI for Analytics

The fear of technical complexity often prevents professionals from exploring AI tools. Many auditors believe they need to learn programming languages like Python or SQL to benefit from AI, but AI tools are becoming increasingly user-friendly. Modern AI systems can generate scripts or perform analyses based on simple queries, eliminating the need for deep technical expertise. Auditors, skilled in asking the right questions, are well-positioned to harness AI’s capabilities. For instance, AI can generate a Python script to identify duplicate invoices, which the auditor can execute in a secure environment. However, similar to how we would never blindly accept an AI-generated RCM, we shouldn’t blindly accept an AI analytics script — the review is still critical.

Myth 4: Using AI Compromises Data Privacy

Data privacy and confidentiality concerns often deter organizations from integrating AI into their analytics workflows. The fear of exposing sensitive information to external AI tools is legitimate, especially given the strict data governance standards in industries like auditing. Plus, most organizations are adopting policies restricting employees from uploading private information to AI like ChatGPT. 

Until off-the-shelf AI develops to a point where we have secure ways to work with sensitive data (which I anticipate will arrive within the next few years), innovative approaches can allow today’s auditors to use AI without compromising data security. For instance, instead of uploading sensitive data, auditors can use AI to generate scripts or tools that run locally on their systems. This ensures that proprietary data never leaves the organization’s secure environment while benefiting from AI’s analytical power.

Myth 5: AI Delivers Perfect Results Without Oversight

A dangerous myth is that AI can autonomously provide flawless insights. While AI is powerful, it is not infallible. It can generate errors or “hallucinations” – results that may seem plausible but are incorrect. Auditors must apply their professional judgment to validate AI-driven outcomes. To perform validation, test results must be confirmed, data integrity verified, and insights must be aligned with organizational realities. AI accelerates the analytical process but doesn’t replace the need for critical human oversight and an auditor’s responsibility to the organization.

The Road Ahead: A Partnership Between AI and Analytics

To answer the CAE’s question from our lunch conversation, analytics is alive and well and continues to be the backbone of data-driven decision-making. As a powerful partner, AI enhances the effectiveness of analytics by automating tedious tasks and uncovering deeper insights. Data governance and auditor judgment remain crucial. 

The future may bring artificial general intelligence (AGI) capable of making judgment calls, but we are still years away from that reality. For auditors and organizations, the key takeaway is this: waiting for AI to reach perfection could leave you behind. Embrace AI tools now to stay ahead of the curve while refining your analytics program in parallel. Those who adapt early will remain relevant and thrive in an AI-enhanced future.