Aahel Iyer
Co-Founder, Midship
March 3, 2024
The audit profession is at a turning point. In recent years, the Big Four accounting firms—Deloitte, PwC, KPMG, and EY—have invested billions in artificial intelligence (AI) to improve audit quality, efficiency, and risk assessment.
While AI-driven audits were once a distant vision, they are now a reality, transforming the way audits are conducted. But what exactly are these firms building? More importantly, what does this mean for mid-market and challenger firms that don’t have billion-dollar tech budgets?
1. Document Review & Analysis (Deloitte)
What They Built: Deloitte’s AI tool Argus automates the extraction of information from documents like contracts, invoices, and financial statements. Traditionally, auditors would spend hours combing through these files, manually checking terms and conditions. Argus does this instantly, flagging potential issues for human review.
Real-World Example: Instead of having an audit team manually review 1,000 lease contracts to ensure compliance with IFRS 16, Deloitte’s AI can scan, categorize, and highlight discrepancies within minutes.
2. Journal Entry Testing & Fraud Detection (PwC)
What They Built: PwC has integrated AI into journal entry testing, one of the most critical steps in financial audits. AI analyzes thousands of entries to detect anomalies that could indicate fraud or accounting errors.
Real-World Example: If an expense is booked to an unusual account or at an odd time (e.g., a large payment processed on a Sunday night), AI flags it for review, helping auditors focus on high-risk areas rather than manually sampling random entries.
3. AI-Powered Risk Assessment & Substantive Testing (KPMG)
What They Built: KPMG's Clara platform uses AI to assess risk levels in audits. It reviews documents, past audit results, and real-time data to help auditors tailor their audit procedures based on actual risk, rather than a one-size-fits-all approach.
Real-World Example: If a company has a history of revenue recognition issues, Clara can flag this and suggest deeper testing in that area, ensuring audits are more focused and effective.
4. Fraud & Anomaly Detection (EY)
What They Built: EY’s AI tools specialize in fraud detection. One of their models was used in the UK to analyze financial statements from ten companies, successfully identifying two cases of fraud that might have gone unnoticed.
Real-World Example: AI can detect subtle patterns in payments that might indicate financial misconduct—like round-dollar payments being made repeatedly to a single vendor just under the threshold that would trigger manual review.
Where Does This Leave Mid-Market and Regional Firms?
The Big Four have the resources to build their own AI tools, integrating automation into their workflows and driving efficiency at scale. Mid-market and regional firms, however, face a different challenge. While they recognize the potential of AI in audit, the question isn’t whether to adopt AI—it’s how.
Firms considering AI have a few options:
Building AI In-House
Some firms might explore developing their own AI capabilities—hiring a team of engineers and data scientists to build custom automation tailored to their audit processes. This approach offers full control and customization, but it also comes with significant investment and complexity:
Cost – Hiring and retaining AI talent is expensive, and ongoing R&D requires sustained funding.
Time – Developing robust AI tools from scratch takes time, often years, before delivering meaningful ROI.
Maintenance – AI models require continuous refinement, especially as regulations evolve and audit needs shift.
For firms with the scale and appetite for internal AI development, this can be a worthwhile investment—but for many, the overhead is substantial.
Leveraging AI Solutions Built for Audit
Mid-market firms are at a pivotal moment. The industry is moving toward AI-powered audit, and firms that take steps now—whether by building or adopting the right tools—will be in a stronger position to scale their practice, improve efficiency, and enhance audit quality.
