Tellen Logo
Header image for How AI Copilots Could Be Deployed to Improve Audit Quality
AI Technology
4 min read

How AI Copilots Could Be Deployed to Improve Audit Quality

October 21, 2024
Ethan Kelemen

How AI Copilots Could Be Deployed to Improve Audit Quality

Recent rises in the rate of inspection findings by the PCAOB have alarmed the governing body and raised concerns across the industry about reversing this troubling trend. As firms struggle to recruit, train, and retain talented auditors, creative solutions are required to maintain audit quality.

One promising solution: AI copilots trained with proper methodology and deployed to assist auditors in real time, offering suggestions to correct common errors before they become deficiencies.

The PCAOB Challenge

In a previous post, we analyzed PCAOB findings trends and identified which firm types were over or underrepresented in deficiencies. Now, we'll explore how AI audit quality assistants could detect and mitigate the trends highlighted by the PCAOB.

Real-World Applications: Learning from PCAOB Findings

Controls Coverage and Risk Assessment

Recent PCAOB findings at PwC highlighted issues with identifying controls related to significant accounts and assertions. During audit planning, ensuring sufficient coverage to support a controls-reliant audit approach can be challenging, particularly for auditors with new or rapidly evolving clients.

The Current Challenge:

  • Evaluating whether in-scope controls mitigate each identified risk
  • Mapping controls to specific transaction subsets and financial statement assertions
  • Time-consuming internal review processes prone to human error

AI Copilot Solution: An AI assistant trained with firm methodology could:

  • Review control-to-risk mapping both quantitatively and qualitatively
  • Assess control descriptions to determine if they actually mitigate identified risks
  • Challenge risk-to-account assertion relationships using audit experience databases
  • Provide insights and suggestions rather than making final determinations

Controls Testing Quality Assurance

Both PwC and KPMG's inspection reports raised issues with controls testing quality. While the testing performance itself is important, the quality assurance component offers significant potential for AI assistance.

Common Errors Identified:

  • Insufficient testing sample sizes
  • Incomplete or incongruent information in testing templates
  • Missing IT system identification for IT-dependent manual controls

AI Copilot Capabilities:

  • Scan entire audit files to review control frequencies, expected vs. actual sample sizes
  • Highlight controls with potential shortfalls in testing coverage
  • Detect inconsistencies in testing templates and documentation
  • Flag missing elements like unidentified IT systems for dependent controls

The Scalability Advantage

Addressing Capacity Constraints

Quality assurance departments are restricted by the capacity of their most experienced professionals. Combined with declining numbers of new graduates entering accounting, the talent to replace and supplement these positions is increasingly difficult to find.

Training Data Evolution

AI quality assistant copilots would be trained on expanding datasets, enabling them to:

  • Evaluate nuanced elements of audit quality over time
  • Learn from historical deficiencies across multiple firms and engagements
  • Adapt to evolving standards and regulatory requirements
  • Provide increasingly sophisticated insights as training data grows

Implementation Strategy

Objective vs. Subjective Reviews

Phase 1: Objective Quality Checks

  • Sample size validation
  • Template completeness verification
  • Required documentation presence
  • Mathematical accuracy verification

Phase 2: Subjective Quality Assessment

  • Control design effectiveness evaluation
  • Risk assessment reasonableness
  • Professional judgment validation
  • Complex relationship analysis

Firm Integration Requirements

Successful deployment requires:

  • Buy-in from leadership on AI quality initiatives
  • Training programs for auditors working with AI copilots
  • Integration with existing audit platforms and workflows
  • Continuous feedback loops to improve AI performance

The Competitive Advantage

Firms that successfully deploy AI copilots for audit quality could:

  • Reduce PCAOB deficiency rates through proactive error detection
  • Improve efficiency of quality assurance processes
  • Scale quality review capabilities beyond human capacity constraints
  • Attract and retain talent by reducing mundane quality review tasks

Looking Forward

As firms integrate learnings from PCAOB inspections into their audits, AI copilots represent a scalable solution to quality challenges. Where these solutions prove effective, the firms that invest in proper training and deployment will be best positioned to reverse recent concerning trends.

The question isn't whether AI will transform audit quality—it's which firms will lead the transformation.

At Tellen, we're developing AI-powered audit quality solutions that learn from industry-wide data to help firms proactively identify and address potential deficiencies before they become PCAOB findings.

Learn more about how Tellen can help improve your audit quality