AI in accounting: No room for complacencyby
While artificial intelligence is highly likely to create efficiencies for accountants, there is no room to cut corners with dataset hygiene and governance.
The recent rise in the adoption of artificial intelligence (AI) continues to prompt new questions and, in some cases, concerns over its role in accountancy.
Given AI’s ability to optimise businesses in functional workflows, spot trends and analyse data, the case for use in the sector is evident. By outsourcing more menial day-to-day functions to automation, accounting professionals can reallocate their resources to more meaningful tasks such as business strategy and relationship management.
However, the key issues concerning the impact of AI on the industry and how it could be adapted to day-to-day practice are far from being resolved. The fundamental problem in this debate centres around how we ensure that AI solutions do not compromise data protection and compliance requirements, both technically and ethically.
In principle, adopting AI should help accounting firms improve their data protection and compliance. The technology has greater capacity than any human to scrutinise data stores and past trends, identify unusual patterns and flag unlawful activity. By using AI correctly, companies should be better positioned to anticipate threats and take preventive measures. However, despite these perceived benefits, key questions remain.
Why is it essential for companies to understand their datasets?
While AI can undoubtedly help alleviate a company’s administrative burden, there remains no room for complacency regarding the oversight and management of data. Companies must remain vigilant regarding their data collection and understand where data is located (locally and across the supply chain), how accessible it is and how the supply chain may be vetted.
“We have seen firms offshoring data on the assumption their outsourcing partner is the processor, only to find out that the data has been passed to a completely different business to process.” Tariq Husain, CEO of GI Outsourcing commented.
It is also important to remain cognisant of the risks associated with data, particularly the reputational damage of data leaks, and how much is necessary to retain.
“Firms must not forget it will be the party highest up the chain that will be accountable for a data breach in the first instance,” added Darren Buckley, director of transformation at GI Outsourcing.
“With data having developed an intrinsic value onto itself, there may be a natural inclination to bolster datasets with ‘just in case’ information, rather than assess each case on its merits,” continued Buckley.
“We recommend that accountancy firms consider the risk-reward ratio of their datasets and the components within them regularly. Undertaking such an exercise may prove an invaluable investment of time for accountants seeking to limit potential risks and demonstrate good governance”.
He added that a critical component of dataset hygiene for companies is to ensure they have the correct policies and procedures in place and that these are backed by practical training.
What are the barriers to the adoption of AI in accountancy?
The resource-saving opportunities of AI adoption are notable in terms of time and financials. However, to arrive at positive outcomes, accountancy firms must invest in the necessary technologies and infrastructure, which, even if done correctly, may come at a cost.
For small businesses, there is a growing sense that the capital cost associated may be a barrier to adoption at this point in the technology’s cycle. As AI products proliferate, it is reasonable to assume costs will come down.
Beyond cost alone, there is a psychological component in that some firms are, by their nature, less likely to embrace technology than others.
Daniel Burrus, considered one of the world’s leading technology forecasters, wrote in August about the associated risks: “It may hurt to hear this, but what works for you now will not work for you in the same fashion a year later, let alone several.
“Technology is always changing, and, as a result, people change. But new technologies play major roles in the merger of individuals’ needs and (the) technology that helps determine what they are. You are very much a part of those paradigm shifts, but as a business leader, you have the power play. You can determine customer needs with ease by implementing said transformative technologies and better serving those customers while also drawing in new ones.”
Regulatory framework and transparency with clients
Regulation and technology operate on a ‘thunder and lightning’ basis, where emergent technologies often come to market before regulation has been implemented, outpacing the governing bodies.
This has been especially true of AI and its rise over the past year, which has left questions surrounding the existing regulatory framework and whether it provides enough guidance for clients and companies alike.
Within this lies the debate over transparency and whether companies have an ethical obligation to appraise their clients on any use of AI. In April, writing on trust in institutions more broadly,
Burrus considered the value of trust and the risk of being too transparent: “Confidentiality is also a trust builder in businesses, organizations, and institutions. There is something to be said for keeping things not as transparent.”
Burrus concluded: “Finding a nice balance between transparency and confidentiality is a future certainty in fostering a trustworthy institution of any kind.”
For accountancy firms, the benefits of AI adoption are significant – but so are the risks of getting it wrong. As an industry, we must upskill to maximise the advantages of these solutions while maintaining the values of trust and authenticity, which have been the cornerstones of our profession for decades.