While many C-level figures will baulk at associated CAPEX, the promise of artificial intelligence in audit is clear. James Dartnell looks at the issues that CFOs need to be aware of when opting for these solutions.Audit automation

Self-driving cars. Facial recognition technology. Innovative drug treatment. For the general public, these are the first things that spring to mind when artificial intelligence (AI) is mentioned. Any old school finance manager or CFO could be forgiven for thinking that these solutions weren’t for them, merely a case of excess CAPEX that will not drive unique value. Nonetheless, while audit may not be able to deliver anything as cutting-edge as life-saving medicine or education services for third-world nations, its automation still stands to bring immense efficiency to an enterprise finance department. High speed, low cost, great agility. All three await those prepared to take the next steps.

There is one benefit of AI in audit that will, on the face of it, have the greatest impact on bottom line. The ability to automate services within a finance department – audit included – will inevitably reduce headcount in finance teams.  On the face of it, this may seem a highly attractive proposition for any CFO. But potential pitfalls are afoot. True, the technology alone making the necessary calculations is extremely powerful – performing trillions of operations per second – but the staff who are tasked with its operations must have the requisite training to ensure that it is not error-prone. Once the necessary algorithms are in place, however, huge time savings are afoot.

Not far behind on the list of plus points is the capacity to reduce risk. Relieving auditing staff of their manual processing duties will give them the power to give greater focus to using their own analytical abilities, and risk falls under this bracket. This is one of the major selling points that must be used when convincing other senior stakeholders of AI’s potential in fraud. AI’s ability to interact with other entities within an organisation’s IT infrastructure could provide the ability to spot, or even predict, fraudulent behaviour. AI has natural language processing capabilities, which means it has similar intuitive and logical capabilities to humans, giving it the capacity to spot irregularities in contracts and documents. On a more basic level, this ability will also bring greater speed in day-to-day contract reviews.

By the same token, while CFOs seeking to reduce operations costs will be keen to make use of AI, they may face an internal struggle to ensure that their audit team remains on side.  Naturally, when audit professionals are told that technology solutions are being brought in to essentially replace the manual aspects of their job, they will feel threatened. Who knows, as AI solutions develop once implemented, there’s always the possibility that its responsibility could extend to the more subjective aspects of a finance department. However the principle of effective training is important here. While for some CFOs this may mean hiring new AI specialists who are skilled in using the software, other CFOs will look to their existing team members to adapt their skillsets for the transition at hand.

One stumbling block, or indeed great springboard for AI – depending on its existing quality – is an enterprise’s IT infrastructure. As the region’s technology infrastructure matures, the adoption of AI solutions in internal audit is bound to increase. While there is undoubtedly room for improvement in terms of SMEs’ IT infrastructures in the Middle East, the region’s larger enterprises are already offering an increasing number of services powered by cloud and Big Data analytics. Technology vendors are also taking steps to bring AI to the region. In November, IBM’s cognitive computing division Watson partnered with Abu Dhabi’s Mubadala in a deal that will see Big Blue’s solutions brought to the Middle East. The exciting effects of Watson have already been witnessed in a host of partnerships around the globe – largely in healthcare and education – and the benefits translate easily into the role of audit.

Examples of effective AI in the region’s enterprises are also not too hard to find. Although not directly involving internal audit, one particular standout use of AI in financial services in the region is Kuwait Credit Bank’s implementation of an automated loan system that can approve a customer’s loan request in just 90 seconds. Having defined an algorithm that interacts with several government entities to verify the applicant’s details, the company is now making vast savings in operations costs from the process, reducing headcount and improving efficiency.

As with any major technology initiative, gaining backing from executive peers will always be an obstacle for a CFO. There will inevitably be figures – even at the c-level – who are skeptical of AI. A number of their concerns will be entirely natural. Although figures are hard to obtain as adoption levels are relatively low, there will always be a fear around the reliability of AI solutions. Although the concept of AI is nothing new, it has so far seen sporadic adoption in the enterprise, and especially in the field of audit. This means that for some, its credibility is lacking. There’s no guarantee that a machine can calculate data with 100 percent accuracy, and so convincing c-level figures must be built around the concept that watertight processes will be in place about those manning AI solutions.

An important part of that process is collaborating with the CIO. Audit may not be the first area that springs to mind as being in need of AI, but again, the promise of greater resilience to risk has to be the main sell in this respect.

Although skepticism will inevitably prevail in the short and medium term, AI in audit must surely be given serious consideration by CFOs and other senior decision-makers in an organisation. As the expectations and role of accountancy and audit change in time, it’s quite easy to envisage a time where AI is not only present, but prevalent in the profession. Ominously, The Economist predicted a 94% likelihood of there being job losses in the accountancy profession over the next two decades. Although surely not an immediate threat for the average Middle Eastern organisation, the threat of Digital Darwinism is probably best not ignored in this regard.