11 May 2023
Close-up of clicking on computer screen

In five years, compliance without AI will be unthinkable

By Kristian N. Stidsen

Leader of Finance Operate & Transformation, Global Compliance & Reporting, Partner, Tax, EY Denmark

Dedicated to developing employees and deliver quality services to EY’s clients. Triathlete and keen cyclist.

11 May 2023
Related topics Tax AI Technology

Many are still cautious about whether AI can be trusted in digital compliance. But soon the idea of compliance will be unthinkable without it.

In brief

  • Real-time compliance is becoming increasingly difficult for risk and control functions to manage due to the accelerating digitization of the world, and the adoption of a data-first approach by global regulators.
  • AI will be an essential tool for achieving data-first, real-time compliance over the next five years, with its supercharged business intelligence, data-driven prediction, and intelligent automation capabilities.
  • To prepare for AI-powered compliance, businesses must ensure data quality, build trust in AI systems and insights through proper governance, control and ethical frameworks, and prepare teams with the necessary skills and responsibilities across finance, tax, legal and IT.

When you stop to think about it, our real-time world is amazing. We move money with a few taps on a phone, communicate with anyone, anywhere, anytime and process huge amounts of data in the blink of an eye. No wonder it’s getting tough for risk and control functions like compliance to keep up.

Data-first, real-time compliance is already on the agenda for many businesses as they respond to the world’s accelerating digitization. Over the next five years, Hans Jessen and I believe AI will become an essential tool to reach this goal. In fact, we’re willing to go further and say that we don’t believe digital compliance can be achieved without it.

Compliance’s digitization challenge

As the world speeds up, lagging data and insights are making it almost impossible for compliance teams to keep up with regulatory obligations and manage the risks of noncompliance.

Global regulators are adding to the pressure as they adopt a digital-first approach focused more on data (inputs) than on returns (outputs). The change is already proving challenging. Our recent tax and finance survey found that 76% of respondents were finding the volume, pace and complexity of global tax reforms difficult to manage.[1] The costs of complying with emerging digital tax administrations (DTAs) is also a concern. Surveyed companies expect to spend $11.1m on an average over the next five years complying with digital filing.

Compliance models are changing in response to this new environment. Compliance is moving upstream, built more proactively into design, operations or the supply chain. And we see growing interest in using technology to achieve real-time or near real-time compliance.

AI will be a compliance essential

Within this context, AI’s supercharged business intelligence will become an essential tool. Data-driven prediction will detect anomalies faster, and arm compliance and business leaders with valuable foresight on high-priority risks. Intelligent automation will be similarly transformative for assessing the impact of rapidly changing regulations and creating regulatory documentation.

Internal fraud offers a good example of AI’s predictive benefits. Because perpetrators often embezzle small amounts sporadically and try to cover their tracks, it’s a particularly difficult problem for companies to detect, predict and control. The cost is also high — expense fraud alone has been estimated to cost companies US$1.9b per year.[2]

AI’s analytic speed and predictive power means it can detect suspicious employee behavior patterns or transactions far more quickly and effectively than non-intelligent systems. It also generates fewer time-consuming false positives, as well as empowering compliance to forecast potential fraud before it happens.

On the automation side, AI’s manual time-saving is substantial. Financial institutions are already using AI to automate legal and compliance documentation, and natural language processing (NLP) is helping other large companies review lease contracts against changing accounting regulations.

AI is primed and ready

While some businesses are exploring AI’s compliance advantages, others may be cautious about whether the technology is ready to be at the center of digital compliance. We believe it is.  

Over the last few years, AI has crossed a threshold in terms of technical maturity, cost and the frameworks required for trusted AI, such as managing bias, transparency and explainability. Major progress has been made in almost all key AI technologies — from machine learning and NLP to speech recognition, image and video generation, planning and decision-making.

In particular, the remarkable power of deep learning models is behind many proven uses of AI. For instance, we know from our EY Tax and EY Assurance work that deep learning and other AI technologies are transforming how businesses classify transaction-level tax data, and how auditors detect accounting fraud.

Preparing for AI’s role in real-time compliance

What does this mean for compliance leaders today? Fundamentally, be ready for AI-powered digital compliance. Make sure your data quality is where it needs to be. While it doesn’t have to be perfect for AI solutions to work (and, in fact, the more advanced the AI, the less perfect it needs to be), better quality data builds trust faster and reduces the time to valuable insights.

Building trust is critical — it’s the foundation for stakeholder confidence in AI systems and insights. There is work to do to achieve it within the finance function: nearly half of finance leaders in our 2020 tax survey did not yet have complete trust in AI system outputs.[3]

To foster trust, start by mapping any new risks that AI brings and use these insights to create the right governance, control and ethical frameworks. As you develop AI solutions, make sure your AI is trained properly and build performance monitoring into ongoing operational processes. You’ll also need to build stakeholders understanding of how AI systems work — in general and within your compliance world.

Finally, ensure your teams are ready. This new model will demand different skills and responsibilities across finance, tax, legal and IT, often within compressed timeframes and with greater regulatory scrutiny. Working together to understand and prepare proactively for AI will position your business to seize the opportunity AI’s disruption of compliance will create.

Summary

The accelerating digitization of the world is making it tough for risk and control functions like compliance to keep up. Compliance teams are struggling to manage the risks of noncompliance and keep up with regulatory obligations due to the lagging data and insights. Data-first, real-time compliance is already on the agenda for many businesses, and AI is set to become an essential tool to achieve it. AI's supercharged business intelligence will detect anomalies faster and arm compliance and business leaders with valuable foresight on high-priority risks. AI’s analytic speed and predictive power means it can detect suspicious employee behavior patterns or transactions far more quickly and effectively than non-intelligent systems.

About this article

By Kristian N. Stidsen

Leader of Finance Operate & Transformation, Global Compliance & Reporting, Partner, Tax, EY Denmark

Dedicated to developing employees and deliver quality services to EY’s clients. Triathlete and keen cyclist.

Related topics Tax AI Technology