ICAEW SE · Chapter 10 · Question 4 of 10
A bank trains an algorithm to assess loan applications using historical lending decisions, which had disadvantaged applicants from certain postcodes. What is the main ethical risk?
Test yourself: pick an answer
Reveal answer & explanation
Correct answer: D) The algorithm may reproduce and reinforce the historical bias in its decisions
Explanation
Data bias occurs when the data used to train or inform a system is unrepresentative or reflects past discrimination. An algorithm trained on biased decisions is likely to repeat those patterns, leading to unfair outcomes. Accountants should challenge how data is selected and captured, and how outputs are tested for bias.
More Data ethics and the impact of technology MCQs
- Q6A company uses a complex artificial intelligence (AI) model to make credit decisions, but cannot explain to rejected applicants why they…
- Q7An accountant uses an AI tool to draft a client's disclosure notes. The draft looks professional but cites an accounting standard…
- Q8What is the main purpose of data protection legislation such as the UK GDPR and the Data Protection Act 2018?
- Q9A trainee pastes a client's unpublished financial forecasts into a free public generative AI tool to help write a summary. Which is the…
- Q10Which of the following is an ethical issue arising from the development and use of new technology in finance functions?