On Creating Accounting Estimates using Machine Learning

The effect of emergent technologies on accountant`s ethical blindness

“…This is primarily because situational factors unconsciously subdue the moral identity, putting more weight on short term benefits than the long-term negative consequences (Aquino et al, 2009). When exposed to ethical dilemmas, humans overestimate their ability to act ethically and fail to realize the power of the situation, resulting in a conflict between the standard of behavior they are expected to comply with and the actual behavior that is committed (Ding et al, 2019). Individuals may fall prey to automatic rationalization and reason that an unethical behavior is for the greater good (Kranacher & Riley, 2019) without taking the time to reflect on the decision making (Laing, 2016).…”

“…In a study comparing Machine Learning (ML) models and managers' estimation of insurance losses, ML models were significantly more accurate than humans with no manipulation of data as analysis of training data is consistent and systematic (Cellan-Jones, 2017). Another benefit of AI is its ability to process unstructured data like audio, video and images and extract insights to support better decisions (Ding et al, 2019).…”

“…ML estimates can thus serve as a first measure for accountants to review. They can also be used as a benchmark against which a case manager can compare human estimates and reconcile if the two are significantly different (Ding et al, 2019).…”

“…With respect to ethical blindness, ML can help accountants avoid several intentional and unintentional fraud activities when preparing financial statements . The latest development in AI capabilities of understanding text, speech and images can strengthen internal controls to support accountant's situational awareness (Ding et al, 2019). Recommendations from AI machines at the time of accounting decision making will help activate ethical standards and minimize the risk of aggressive accounting.…”