Skip to content

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

A bank pilots a generative AI assistant to draft summaries of regulatory filings and client communications. Which risk is most specific to generative AI in this use case compared with traditional statistical models?

Hallucination is the risk most specific to generative AI: the model can generate confident, fluent text that is factually wrong. In drafting filings or client communications this creates operational, compliance and conduct risk, requiring human review and controls, unlike traditional statistical models.

  1. AHallucination, where the model produces fluent but factually incorrect contentCorrect
  2. BHeteroskedasticity in the regression residuals
  3. CBasis risk between the hedge and the underlying asset
  4. DConvexity risk from embedded prepayment options

Explanation

Generative models can produce plausible but false outputs, a risk specific to them. Heteroskedasticity, basis risk and convexity are issues for statistical, hedging and fixed income contexts respectively, not distinctive to generative AI.

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