Auditors are Innovators: Using AI to Realize the Future of Audit

• Jon Fritsch
• Oct 24, 2024
• 5 min read

When you think of auditors, the first words that come to mind might be "compliance," "risk management," or even "traditional." But what if I told you auditors had evolved to become an unexpected innovator in the business world?

At Ally, our internal audit team is breaking stereotypes and embracing generative artificial intelligence (gen AI) to not only redesign our processes but redefine our role and realize the future of audit. In strong partnership with Ally’s Technology organization, we’ve developed a generative AI-enabled application that’s not just enhancing our current practices but also helping us peer into the future of the audit profession. While Ally is still early days of its AI journey, by thinking big and acting small, Audit is positioning our team for the future by being ready to navigate and lead in this dynamic environment.

The Use Case:

While we envision a future where gen AI is integrated throughout our audit processes across risk assessment, annual planning, audit execution, and talent management, our first use case focused on audit planning. By starting with audit planning as our first use case, we could see the potential to save significant time while also not including customer or sensitive data, which is common in fieldwork. Our new application is designed to optimize the planning and scoping phases of audits. By quickly generating relevant content on potential risks, expected controls, and recommended test procedures, while seamlessly integrating human-in-the-loop for review and editing, this tool has the potential to revolutionize how we approach our work.

The development process was a collaborative effort, involving cross-functional teams of critical technology partners, architects, developers, and audit professionals. The hypothesis was clear: could gen AI streamline our audit processes without compromising quality? The answer was a resounding yes.

During the pilot phase, the generative AI model processed basic project details based on prompts developed by the audit team. The output was comprehensive risk and control matrices (RACM). The results were impressive, with gen AI model identifying potential risks generally consistent with current practices and suggesting controls that aligned with industry best practices.

Use Case Results:

The pilot phase yielded promising results, demonstrating the viability of the approach. Here are several key results:

• Time Efficiency: We estimate time savings of more than 60 hours per audit or roughly 2,400 hours annually. A RACM can take up to two weeks to manually draft for management review; in minutes, the AI-generated RACM is developed and ready for auditor review and editing, then management review.

• Output Accuracy: Over 75% of the AI-generated RACM was rated Good or Great by the Audit team when assessed on reasonableness, methodology adherence, and prompt formatting.

These early results for just one aspect of our process underscore the potentially transformative power of gen AI in our work. Ally Audit currently has several other use cases in development and a roadmap for future work informed by continuous feedback and input by Audit users. The work we’ve done to explore how best to integrate gen AI into our processes has unlocked opportunities to redesign and streamline our methodologies. By embracing an “Art of the Possible” mentality, we’ve been able to challenge our current state and envision what could be. Now more than ever, we’re looking at the future of our function and how to support Ally’s business as gen AI and other emerging technologies are explored.

Challenges and Key Learnings:

The journey has not been without its share of learnings. The cycle of prompt writing and testing with foundational LLMs can be tedious, and, at times, head-scratching – all the more reason to keep a human in the loop of all our interactions with gen AI models. User adoption and change management, more generally, takes time and a concerted effort to do well. Small and proficient teams can move quickly to build solutions, but it takes a community approach to change hearts and minds and processes.

The main learning is the power of strong partnership and collaboration between our internal audit and our technology teams. This collaboration, built from the top with our Chief Audit Executive Stephanie Richard and Chief Information, Data and Digital Officer Sathish Muthukrishnan, has been crucial in realizing our vision and developing an application that meets our needs while exploring what’s possible. Over the past several years, Tech and Audit have become partners – way beyond a typical vendor/client relationship, even within an organization. They have been by our side in evaluating, developing, testing, troubleshooting and learning. We’re thankful for that collaborative spirit.

Leaping Toward the Future:

For Ally Audit, the launch of this gen AI application is a strategic leap towards a future where audits are more efficient, proactive, predictive, and – ultimately – more value-added. By integrating gen AI into our processes, we’re not only enhancing current operations but also positioning ourselves at the forefront of the evolving role of internal audit.

The emergence of gen AI both coincides with and accelerates a critical juncture for internal audit. Industry standards are evolving to address the changing role of internal audit and the impact of emerging technologies. Internal audit industry groups have set out an ambitious vision for the profession, with gen AI and machine learning at the forefront paired with more collaborative and integrated stakeholder relationships. Real-time advisory services are increasing in prominence, and, while we don’t view things this way at Ally, internal audit is seeking to be viewed more as a valued partner and advisor.

Ally is meeting the moment with caution, but also optimism. The potential use of this application and approach by other testing and assurance teams within the organization further positions this gen AI application as a valuable enterprise asset. Moreover, our experience in developing AI solutions better equips us to audit the broader organization's use of both traditional and generative AI, helping us remain at the leading edge of technological innovation. We can meet our core charge to provide assurance on risk management, governance, and controls as more use cases emerge and new technologies are introduced.

As we look ahead, the vision of audit enabled with AI becomes ever more tangible. For the internal audit function, gen AI is a tool and a catalyst for change, helping us work towards a vision of the future where audits are more efficient, insightful, and impactful. And at the heart of this transformation are our auditors as innovators, working hand-in-hand with our technology team to turn vision into reality. By thoughtfully embracing AI, we’re fostering a culture of continuous improvement and inspiring an innovative mindset within our audit team. Our journey exploring generative AI has enabled us to see the future of audit more clearly and take steps down that path – and step into our role as innovators.