Free · 5 scenarios · About 8 minutes

The AI Judgment Challenge.

AI recommendations can sound confident and fluent even when they are wrong. This challenge puts you in five fictional financial-services situations and asks one question each time: is the AI right, or is it leading you?


01 — How it works

Read. Decide. See your score.

Step 01

Read the scenario

Each one includes the case details and the AI's recommendation, just as you'd see it at work.

Step 02

Make your call

Accept, challenge, or reject and escalate. There's no time pressure, so take the pause judgment needs.

Step 03

Get your result

See your score straight away, with feedback on where your judgment held and where the AI took the lead.


02 — The three calls

Every scenario, one decision.

Accept

It holds up

The recommendation is sound, so you proceed as the AI suggests.

Challenge

Something's off

Something does not add up, so you query it or ask for more information before acting.

Reject & escalate

Stop and raise it

The recommendation is wrong or breaches policy, so you stop it and refer it to a senior reviewer.


03 — Take the challenge

Your five scenarios.

All scenarios are fictional. This is a practice tool with an indicative score, not a certification or a professional assessment.

Start the challenge

Opens the five scenarios full-screen in a new tab. Your result and the Practice Pack link appear at the end, and this page stays open in your other tab.

Finished? See the Practice Pack ↓

Form not loading? Open the challenge in a new tab →

04 — Go further · Early access

The AI Judgment Practice Pack.

Practise knowing when to accept, challenge or stop a confident AI recommendation. Thirty realistic scenarios where the AI recommendation is confident, plausible and sometimes wrong.

  • 30 scenarios across lending, insurance claims, fraud, superannuation member services, hardship and risk
  • The right answer and the reasoning behind every scenario
  • A breakdown of the errors you missed and the sound recommendations you unnecessarily challenged
  • Unlimited retakes, in any web browser, with nothing to install
Early-access preorder
A$79+ applicable taxes

Delivered by email on Friday 4 December 2026, with your link and access password. If it isn't delivered by that date, you get a full refund.

Secure checkout via Gumroad. All scenarios are fictional. A practice tool, not a certification.

Practise together as a team

Use the 30 scenarios with up to 10 named users in one organisation, with permission to discuss the cases and explanations in an internal team session.

Team access is quoted separately. The A$79 preorder is for individual use.

05 — Questions

Good questions.

Who is the challenge for?

Anyone who reviews, approves or acts on AI-supported recommendations: lending, claims, fraud, member services, risk, audit and the leaders accountable for those decisions.

Do I need to work in financial services?

No. The scenarios are set in financial services because the stakes are clear, and each one gives you practice in checking an AI recommendation against the facts and policy in front of you. The habit of checking may be useful in other settings, but this challenge does not test that.

Is this a test of whether I'm capable?

No. It's practice. It gives you five practice cases where an AI recommendation sounds certain, with an indicative score and feedback on each answer.

When will I receive the Practice Pack?

On Friday 4 December 2026, by email, with your link and access password. If it isn't delivered by then, you get a full refund.