Please make sure you are using a supported browser. To find out more click here.

Looking Beyond the Digital Accounting Horizon: Minding the Gaps and Avoiding Decision Drifts

By Fearghal McHugh and Dr. Trevor Clohessy

As there are new working models which offer new and exciting opportunities for business, employees, society; the significant and strategically important technologies and policy that will affect opportunities are around Distributed Ledger Technology (DLT) and Artificial Intelligence, operating within business strategy and government policy.

DLT records transactions between entities in blocks. The data in each block is secure, distributed, and anonymised. DLT technology was first used to secure the Bitcoin cryptocurrency and came to global renown in 2015 when the technology featured in the Economist magazine and once again at the end of 2017 when Bitcoin reached a peak value of $20,000. DLT in this context is referred to as blockchain. However, whereas Bitcoin and blockchain can be described as the first use case of DLT, it is now being used across multiple industries in various guises. We argue that DLT needs to be separated from blockchain as in the way it will be implemented and how it will affect the way business will be done; potentially removing the need for some businesses for the customer. Too often blockchain is associated with the negative connotations of cryptocurrencies, initial coin offerings, and so DLT as the backbone should be the focus.

Companies that adopt DLT can derive business supply chain benefits such as reduced operations and transactions costs, improved accountability and traceability, enhanced security, and improved decision making because of real-time data and information. One possible disadvantage of DLT is the removal of trusted intermediaries which are replaced by decision consensus in a peer to peer network. Furthermore, DLT governance structures such as private permissioned DLTs are not truly decentralised and thus the benefits described earlier cannot be accrued. Core to DLT functionality is the concept of smart contracts which represent digital manifestations of traditional contracts. However, unlike the former, smart contracts execute the terms and conditions automatically without any human intervention. This form of AI can result in a multitude of benefits and is being embraced rapidly to alter the customers touch points with organisation which can result in a distance between the customer and the business. AI can also have disadvantages such as biases or the execution of decisions automatically which can be underpinned by errors.

Hence there is now policy being developed in relation to AI. The European commission has delivered its first ever Artificial Intelligence (AI) regulatory framework that will protect the safety and fundamental rights of citizens from the decisions made by high-risk AI information technology (IT) systems; imposing requirements on businesses that are using or who intend to use high-risk AI based decision systems. The proposed legal framework applies to the “use of AI with its specific characteristics (e.g., opacity, complexity, dependency on data, autonomous behaviour) can adversely affect a number of fundamental rights so as to foster the creation of ‘Trustworthy AI’ to eliminate the occurrence of biased AI based decision-making, involved in employment and other areas which can infringe on fundamental human rights. the new AI legal framework will create trust, transparency, and accountability between public authorities, businesses, and citizens.

It is critical to observe the implementation of these; the systems will fundamentally change working practice, process, the working environment and income streams. Technology’s impact on the accountant role within businesses will move from the tangible transactional level to focus on the intangibly gaps and also result in the potential generation of  decision drift that will need to be carefully monitored by all in business.

The core concern arises when we begin to use and more importantly rely on these systems to do business. This reliance is the issue, as it may remove many of the current operational processes, that enable the experiential gathering of skills to make decisions and move this ability into the systems. This reliance maybe shifted on to the systems of DLT and AI rather than the person who used to be involved; this could ultimately lead to generic systems decisions with bias and decision drift.

Decision drift may develop from, the user relying on the systems to make decisions and the systems appearing reliable to users, in addition to people being more physically distributed, via newer working environments. This may result in an authority being delegated to the decision systems over time, leading to a drift, like strategic drift. So operations appear to be going well and are then left to run as is; when in fact the market has moved but the organisation has not.

To ensure that this is not going to happen, we need to first identify the intangible gaps, where we can input and evolve ourselves and the business, and where systems do not have the intuition that we have to ensure that organisations are going the right direction. Intangible gaps are areas such as culture, communications, empowering people to make decision, experiential training, market behaviour (not based on trends analysis but in people observations and gut feeling about the market build on experience). An intangible gap analysis is needed to ensure these gaps are filled by rounded experience and not just system experience but to work in conjunction with the DLT and analytics. This should ensure that systems are not allowed to create a blind feeling of reliance on systems to make the decision. Outsourcing, even the routine decision making is a convenient option but has consequences.

Consequences can be seen in bias that are built into systems, such as advertising systems used by Facebook, recruiting systems used by Amazon, Health care systems and criminal system from the US government. These systems resulted in a form of bias against female selection for roles (recruitment systems); towards genders selected in a traditional manner for specific roles (advertising systems) white patients favoured over others (health systems) and African Americans identified over others (criminal systems).

Another step is to update the skills required to deal with all the technology change alone. An Irish blockchain industry technology report outlined how the core competencies and skills required for these technologies the following categories:

  • foundational technology (e.g., cryptography, public key architecture);
  • Distributed ledger technology (e.g., mining, consensus algorithms);
  • Forensics and law enforcement (e.g., money laundering, darknet);
  • Markets, economics, and finance (e.g., business modelling, cryptonomics);
  • Industrial design (e.g., supply chain, Internet of Things) and
  • Regulations and standards (e.g., smart contracts, and governance frameworks).

From an accounting perspective, certain traditional skill set elements relating to accountancy practices will be eliminated or reduced (e.g., reconciliations, provenance assurance). A report by McHugh and Clohessy highlighted how blockchain and DLT transactions will enable new value-adding activities (e.g., ability to check transactions in real-time). However, while the range of extant skills required will change, this change need not be byzantine. It is envisaged that skill categories, markets and regulations outlined above will be important for bridging the DLT literacy gap between various business and technology stakeholders.

Looking forward, accounting practices can examine their business models to derive value from DLT and AI via a intangibility gap analysis and monitoring the development of decision drift; as there will be an end to traditional ways of doing things and a new business approach which will be divisive, pervasive, and transformational all at the same time and will encourage accounting professionals to look ahead and not base most operations and decision making on past data. Present and predictive transacting data systems with transparency and integrity inbuilt is a DLT‑AI future; we must be conscious that we use our experiential judgement and remain the decision maker as we develop towards being the digital accountant, minding the gaps and avoiding drift.


About the authors:

Fearghal McHugh is a lecturer in the Department of Enterprise and Technology, at Galway-Mayo Institute of Technology School of Business since September 2010. Learning interests are around decision making in strategy, governance, ethics, and the impact Information Systems have on process design and the system outputs that affect decisions across an organisation. Fearghal has also written articles in the areas of knowledge management, AI, and blockchain for accountants and business.




Dr. Trevor Clohessy is a lecturer and researcher in the Department of Enterprise and Technology, at Galway-Mayo Institute of Technology School of Business since September 2018. His research interests include blockchain, artificial intelligence, and digital transformative technologies. Trevor has written a book called “Digital Identity Politics”, which explores how new technological developments such as Artificial Intelligence and social media analytics have impacted political campaigns. 

Connect. #CIMAIreland

Do you want to be in the know? Follow CIMA Ireland social networks to find about important events and local news.