Gale's Content


Artificial Intelligence is the Real Thing

Reprinted with permission from Accounting Today

In late 2015, Eli Fathi was named CEO of the artificial intelligence firm MindBridge. Fathi’s enthusiasm for AI was sparked by the purchase of DeepMind by Google in 2014, demonstrating the potential power and opportunities for AI-based applications.

Fathi, a Canadian serial entrepreneur, had headed several tech startups when MindBridge founder Solon Angel tapped him for the new venture. Timing was excellent, as Fathi had been pursuing the potential of applying artificial intelligence (AI) and deep learning to financial data.

In a recent interview with Eli, I asked what led to his interest in AI in accounting. Fathi described research revealing that, worldwide, the gross domestic product  (GDP) was losing about five percent of all financial data due to fraud, mismanagement, errors, bribes, etc. Today, GDP is at $85 trillion and still losing five percent, with only a small percentage of the loss being detected by current analytic tools. The lion’s share is caught through whistleblower tips.

If AI could capture the loss, who is in the best position to do so? Answer: auditors, because they are on the front-line identifying unusual transactions and misstatements in financial records.

Other Challenges

One key obstacle slowing adoption involves the traditional approach toward finding anomalies—sampling. When auditing standards were first established, no system existed to permit testing of large data sets. As a result, sampling was best practice.  Today however, though technology permits us to analyze whole data sets to identify unusual transactions, ISA 240 and other regulations still require sampling.

It’s a kerfuffle and a conundrum rolled into one. Sampling supports the spirit of the standard, but not the intent, which is to seek inconsistencies. As such, the standards support a lower level of scrutiny than today’s technology enables. This leaves auditors hamstrung by regulations that have not caught up with technology, limiting their ability to deliver more value to their clients.

Because sampling is all that’s required, the information presented within the audit opinion using the existing tools and methodologies may not always provide optimal results. Using AI-based tools will enable auditors to overcome the deficiencies of sampling and rules-based tools to provide more comprehensive service and value to their clients.

Educating clients on the value implications is key, Fathi emphasizes. AI moves businesses from hindsight to insight to foresight, unearthing predictive capacity. But that means auditors must discover the applicability of this powerful tool to their unique clients. The first iPhone offers a comparable example. It was only once the sleek white boxes were opened and consumers experienced the revolutionary device that the vast potential of the iPhone was recognized.

In working with firms, I’ve discovered the most compelling technology use cases are industry specific. Yes, AI can be valuable for any sector, but it’s the discovery of industry-specific nuances that create value. Comparably, the iPhone is only as valuable as the apps that fit the user’s specific needs.

According to Eli, getting to this future promise presents another challenge, the change in the role of the auditor. Where in the past the auditor has been in a compliance role, tomorrow’s auditor will be in a position to provide additional high-value services, both within and beyond the audit.

I find that while technology continues to deliver ever more capability, managing partners can and should work to ready their firms for what’s next. The first task, after reassessing the role of the auditor, is to nurture product management skills among his leaders. This is the functional discipline of developing new services for different markets, a little-known specialty among firms.

Eli added that an important component is the human behavior change needed throughout the firm , which is just as important as the adoption of technology (think Airbnb or Netflix).

Tips for MPs

I asked Fathi how managing partners can get a deeper understanding of AI and work toward building adoption. He advises MPs to:

  • Champion change. MPs will need to embrace the change from the top, spread the word and provide resources.
  • Understand that younger members and emerging leaders are eager for new tools that may not appear critical to all members of the firm.
  • With AI moving through the typical technology lifecycle of high expectations and relatively low adoption, MPs need to know that there is light at the end of the tunnel. This is especially true for early adopters who will realize considerable value by being early to market with a significant competitive differentiator.
  • Realize that AI will enable you to leverage the power of your people. The higher-level work will make it easier to attract and retain them.

As we wrapped up the interview, Eli summarized that the twists and turns of progress are hard to predict, and even harder to navigate. For firms committed to pivot from compliance to consulting, AI is an essential strategic tool. And while artificial intelligence will not replace auditors, those auditors who know and use it will replace those who do not.