Free assessment
Eighteen questions, about 7 minutes, and a straight answer about where your practical AI skills actually sit. No account needed to start.
Free · 18 questions · about 7 minutes
What it covers
The same five areas the certification is built around, tested at a lighter level. This is about how you use AI at work, not how models are built.
How language models produce output, and the limitations that follow from it. Covers next-token generation, training cutoffs, hallucination, inherited bias, and the effect of settings such as temperature on the character of a response.
Giving a model the instruction and the material it needs to do the job. Covers specificity, worked examples, supplying source documents, structuring output, and managing a conversation as context accumulates.
Systems where a model calls tools or takes multi-step action toward a goal. Covers what distinguishes an agent from a single exchange, what a tool call is, where autonomy should stop, and how a human stays in control of the loop.
Deciding whether output is good enough to use, and knowing what to do when it is not. Covers checking claims against sources, calibrating scrutiny to stakes, spotting confident errors, and retaining accountability for work a model helped produce.
Protecting information and people when using AI at work. Covers what may be entered into a tool, organisational policy and approval, personal and confidential data, third-party obligations, and recognising a situation that should be escalated rather than handled.
What you get
Your result places you on one of four levels, with a breakdown by area and the two places worth your attention first. We ask for an email at the end to unlock the full report.
0–39%
Knows what these tools are and has used them, but not yet in a way that holds up under scrutiny at work.
40–69%
Uses AI regularly and gets useful results, with gaps in verification, context handling, or data judgment.
70–89%
Works with AI deliberately and checks it properly, at the level the certification is being written to.
90–100%
Fluent across prompting, delegation, verification, and data judgment. Able to set practice for others.
This is a free self-assessment, not the certification exam and not a credential. It is unproctored and its result is not verified. It is here to tell you where to put your effort.
