Skip to content

FRM Part II · FRM Exam Part II · The Financial Stability Implications of Artificial Intelligence

Which statement best describes 'adversarial' attacks on a bank's AI-based fraud detection model?

Adversarial attacks involve deliberately crafted inputs or manipulated training data that cause an AI model to misclassify fraudulent activity as legitimate, evading detection. They exploit weaknesses in the model itself, unlike ordinary regulatory actions or voluntary data sharing.

  1. AInputs are deliberately crafted or training data is manipulated so the model misclassifies fraudulent activity as legitimateCorrect
  2. BThe model is shut down by regulators for exceeding its capital allocation
  3. CCustomers voluntarily share data to improve model accuracy
  4. DThe model's outputs are always more accurate when exposed to unusual data

Explanation

Adversarial attacks, including data poisoning and evasion inputs, exploit model weaknesses so malicious activity is misclassified as legitimate. The other options describe unrelated or incorrect situations.

Did you get it right without looking?

One question tells you little. A timed set on The Financial Stability Implications of Artificial Intelligence shows your real accuracy, how long you take and where you lose marks.

Practise The Financial Stability Implications of Artificial IntelligenceFree diagnostic mock

More The Financial Stability Implications of Artificial Intelligence questions