Lightbulb Moments in Data Analytics and AI

We all have lightbulb moments where a concept clicks, especially when we are shown how new technology can improve our lives. Nearly every big technological leap over the past 30 years made an older one less significant. Of course, we face a natural human resistance to change with every advancement, and some people cling to old habits. 

We have this same situation today in internal audit. Using data analytics is easier than ever with tools like artificial intelligence (AI), ChatGPT, and AuditBoard’s built-in data analytics and AI capabilities. Still, many internal auditors are hanging on to old ways of working. We share a conviction that embracing technology like data analytics and AI is the key to success in internal audit, so we wanted to share our own lightbulb moments. 

Data Analytics Lightbulb Moments

Data analytics has proven itself a key capacity multiplier for internal audit teams, enabling us to make use of the incredible wealth of data now available from both internal and external sources. For both of us, our data analytics revelations occurred when we recognized the significance of the possibility to test full populations rather than taking a sample, which in some cases would represent less than 1% of the population we’re testing. 

Richard: I had a major lightbulb moment when I was the Deputy IG of the Postal Service. We had launched a money order service for people working in the US to send money back to loved ones in other countries. One senator was concerned the service could be used for money laundering, so he requested an audit to look for unusual transactions. We had just seven days to go through hundreds of thousands of transactions – an impossible task using traditional testing methods. I took the problem to our technology auditors. Three days later, they came back with a completed audit. A team of 4 people reviewed over 375,000 transactions and pinpointed a very small number of irregularities using early data analytics software. That was my “aha!” moment, and I have been sold on the power of data analytics and technology ever since. 

Trent: All data analysts I talk to have an “aha” moment. When I was in public accounting, we were testing to better ensure terminated users had been removed from an application. The senior manager said to use a VLOOKUP in Excel to compare the general ledger user list to a terminated employee list from human resources. The test took 30 seconds, and we had tested the full population of thousands of users. We were able to identify five terminated users that still had access. From that day, I showed every staff member how to do a VLOOKUP on day one. Even basic Excel functionality can be a powerful analytics tool, and the moment auditors see that power in action, it is a gateway into more advanced testing techniques.

Artificial Intelligence Lightbulb Moments

As we step into a new age of technology, auditors are having lightbulb moments with AI and tools like ChatGPT. We’ve found that when people encounter an actual AI use case, a lightbulb goes on, and we often hear people say, “I didn’t know it was so easy.” 

Richard: I realized the potential that generative AI presented almost immediately upon launch of ChatGPT in 2022. But the real lightbulb moment for me was when a CAE showed me how her audit team was using it to jumpstart their risk assessment process. ChatGPT provided a comprehensive, insightful list of current and emerging risks tailored to the industries and geographic regions her company operated in. They were essentially using AI as a research assistant, but what AI provided in minutes would have taken days or weeks in the past.

Trent: One of the most practical AI tools is the ability to summarize meetings in Teams. I was recently working with a partner at a Big Four accounting firm who had a large ‘All Hands meeting’ coming up. I showed her how to use the AI in Teams to summarize meetings from a generated transcript, and that was her ‘aha’ moment. The ease of use of these tools reassured her, and now she has her entire team using this AI and sharing meeting summaries to improve communication.  

I still have lightbulb moments, and recently experienced one using ChatGPT. One of the many things you can use AI like ChatGPT to do is write analytics scripts. Someone mentioned using an advanced algorithm (i.e., forest isolation) to identify outliers and anomalies in the general ledger. I figured I could do this independently, but it would take me about two weeks to get it right. I tested this idea by removing Python, my analytics tool of choice for this solution, from my computer and prompting the AI to tell me what to do from start (i.e, installing Python on my computer) to finish (i.e., implementing the model analyzing the results). My jaw literally dropped as the AI response walked through everything I needed — I had this powerful new technique ready to use in under an hour!

Find Your Lightbulb Moment 

New technology can be intimidating at first, but it quickly becomes a part of your regular life. Remember, you’re not going to break anything. If something doesn’t work the way you want, close it, open it again, and now it’s fixed. There’s no point in being a late adopter; that’s just setting yourself up to be left behind. Richard’s advice is to be courageous and adventurous. Try new tools as soon as you can. This will not only keep you ahead of the curve but also motivate you to experience your own lightbulb moments.