Seven technology characteristics posing an ethical challenge

Emerging technologies such as artificial intelligence (AI) are changing the way accountants work at a rapid pace. These technologies are advancing month-to-month and are being adopted for increasingly more tasks in accounting firms and finance teams.

But these technologies come with risks that need to be reassessed on a regular basis, forcing accountants to consider the ethical impacts of technology use.

The International Ethics Standards Board for Accountants (IESBA) has attempted to address this with its latest guidance on the topic.

It identified several characteristics of technology that potentially pose an ethical challenge for accountants.

Opacity: this may limit transparency and explainability, which would make it more difficult to spot biases, errors and assumptions built into the system.

Non-determinism: similar or identical inputs may create different outputs.

Data dependence: data quality, completeness and relevance will influence the quality of outputs.

Perpetual adaptivity: enabling systems to modify behaviour or outputs in response to various inputs, feedback or changing conditions.

Autonomy: the ability for systems to operate independently, with little human oversight.

Scalability: the ability to maintain performance across larger volumes of data or tasks. This may amplify existing biases or errors.

Speed: the ability to process or perform tasks and generate outputs in a short timeframe. This could increase the scale and impact of both benefits and errors, and could make it harder to implement effective human oversight.

Exercising judgement

Accountants are required to have “an inquiring mind”, exercise professional judgment and use the reasonable and informed third-party test to assess whether the use of technology threatens compliance with the fundamental ethical principles. Ethical considerations should apply throughout the lifestyle of any technology, from design and procurement to replacement and disposal.

Technologies often work within wider systems and in tandem with other tools. This could create or boost the ethical threats associated with that technology, and accountants need to consider these potential effects when assessing the ethical threats of the technology they are using.

“For example, AI and machine learning may be combined with data analytics, cloud computing, robotic process automation, or blockchain technologies within a single solution, resulting in increased complexity, opacity, and potential impacts on transparency, accountability, privacy and decision making.”

IESBA recommends a holistic approach to ethical evaluations, considering how all technologies interact within an ecosystem.

Constant monitoring

Under the Code of Ethics, accountants are required to re-evaluate the ethical implications of technology and how to address any threats that may have arisen. For example, the impact of ‘model drift’ – a decline in a model’s predictive accuracy – or updates to software or algorithms that are implemented by the technology supplier without full transparency.

Maintaining professional competence and due care when technology changes so rapidly requires a continuous learning approach, according to the guidance. “A PA is expected to have, or obtain, a sufficient understanding of the technology to evaluate whether its use is appropriate and whether reliance may be placed on its outputs.”

Appropriate guardrails and controls are required to manage and mitigate threats; IESBA gives the example of a policy prohibiting the use of unvetted or unapproved software.

Accountants need to also be aware of the fraud risks posed by AI. Deepfakes, synthetic data and other generated content can make fraud attempts more convincing. Finally, accountants need to monitor the regulatory environment around technology use to ensure that they remain compliant.