Skip to content
calwm

Red pen: Guidelines for responsible AI

Exam answers on Guidelines for responsible AI are often almost right. Each round hands you a short document with a few mistakes hidden in it. Mark the faulty lines, say why each one is wrong, and see what a good version would say.

Find the mistake before the auditor does.

Grab the red pen2 documents, about 4 min

What you will practise

You will practise Guidelines for responsible AI, with questions touching AI transparency and responsible AI practices.

  • AI transparency
  • Responsible AI practices

It is part of the AWS Certified AI Practitioner track (AIF-C01), where Guidelines for responsible AI is about 14% of the exam.

A taste of what you will get

  • Theo noticed that strong candidates over 50 rarely make the shortlist.

    Sample document 1: Ben's fairness plan for the shortlisting model. Which lines are wrong, and why?

    • Goal: shortlist the 20 strongest CVs for each role
    • Training data: last year's hires, almost all under 40
    • Success check: overall accuracy only, currently 91%
    • Recruiters review every shortlist before candidates hear back
    • Record the model's intended use and limits in a SageMaker Model Card

    Answer: Line 2 (Training data: last year's hires, almost all under 40): Few older examples, so it learns to judge them poorly. Better: Use balanced, representative data, and check how many examples each age group has. Line 3 (Success check: overall accuracy only, currently 91%): A good overall score can hide a group treated worse. Better: Compare shortlisting rates and errors for each age group (subgroup analysis).

  • The regulator may ask how any shortlisting decision was made.

    Sample document 2: Ines's transparency checklist. Which lines are wrong, and why?

    • Candidates aren't told that AI screens their CVs
    • Each model's intended use and limits are recorded in a Model Card
    • Any model is fine for the regulator, as long as it's accurate
    • Recruiters can override any shortlist decision
    • Candidates can give feedback on the process

    Answer: Line 1 (Candidates aren't told that AI screens their CVs): People should know, so they can make informed choices. Better: Tell candidates that AI helps screen CVs, and how to ask for a person to review. Line 3 (Any model is fine for the regulator, as long as it's accurate): Accuracy alone can't explain why a candidate was rejected. Better: Prefer a model whose decisions can be explained, and record why it was chosen.

Advertise here

How scoring works

  • Every correct answer pays 10 XP for each difficulty level, so an easy item is worth 10, a medium one 20 and a hard one 30.
  • Answer with time to spare and you can earn up to 5 bonus XP on top.
  • A wrong answer pays nothing. It does not cost you anything either.
  • Flashcards pay 2 XP each time you grade yourself 3 or higher.
  • Using a hint costs 2 XP, so use it when it helps, not when it is comfortable.
  • Finish a session on consecutive days to build a streak. Your progress lives in your browser, and signing in syncs it.

Quick answers

How long does the red pen on Guidelines for responsible AI take?
About 4 minutes for 2 documents. Pause whenever you like.
Do I need an account to play the red pen?
No. Progress is saved in your browser, and signing in syncs it across devices.