Skip to content

ACCA Strategic Professional · Advanced Performance Management · Technology and information systems

Brightwell Bank uses a machine learning model to score loan applicants, and the score feeds into a sales team's bonus scheme. An internal review finds the model rejects applicants from certain postcodes at a much higher rate, because past lending data reflected historic bias. Which is the most appropriate management accounting response?

The best response is to test the training data and model outputs for bias and to review the bonus scheme so it does not reward biased outcomes. Machine learning learns from past data, so historic bias can be reproduced; objectivity cannot be assumed, and incentives must support ethical, fair decisions.

  1. ATest the training data and model outputs for bias, and review the bonus scheme so it does not reward biased outcomesCorrect
  2. BKeep the model unchanged because algorithms are objective by nature
  3. CRemove all historical data and rely on managers' unaided judgement
  4. DIncrease the bonus weighting on approvals to offset rejections

Explanation

Models trained on biased historical data reproduce that bias, so the data and outputs must be tested and governed. The reward scheme should also be reviewed so it does not encourage unethical or discriminatory behaviour. Assuming objectivity ignores the risk, and raising approval bonuses worsens incentives.

Did you get it right without looking?

One question tells you little. A timed set on Technology and information systems shows your real accuracy, how long you take and where you lose marks.

Practise Technology and information systemsFree diagnostic mock

More Technology and information systems questions