The EU AI Act Readiness Scorecard.

20 questions across Articles 9, 10, 12, 13, 14, 15 — risk management, data governance, record-keeping, transparency, human oversight, and cybersecurity. Every "no" is a gap.

Free PDF, no calls, no spam.

Six articles, twenty questions

The same high-risk obligations a conformity assessment walks before agents touch production. Score yourself before the assessment does it for you.

Risk management (Art. 9)

Do you run a risk management system across the agent lifecycle — identifying, evaluating, and mitigating foreseeable risks, with a register your conformity assessment can read from?

Read Article 9

Data governance (Art. 10)

Is grounding and fine-tuning data tracked with source, version, and EU residency — documented lineage showing exactly what informed each agent?

Read Article 10

Record-keeping (Art. 12)

Does every agent action write to an automatic, tamper-evident log over the system's lifetime — or does traceability depend on someone remembering to switch logging on?

Read Article 12

Transparency (Art. 13)

Does each agent ship with documentation a deployer can interpret — what model backs it, what it can and cannot do, and how its output should be read?

Read Article 13

Human oversight (Art. 14)

Can a natural person intervene, override, or stop the agent mid-task — oversight as a configured control, not a manual habit?

Read Article 14

Accuracy + cybersecurity (Art. 15)

Are agents evaluated before changes ship, run behind prompt-injection defences, and held to scoped tool permissions so an adversarial input can't reach a capability you never granted?

Read Article 15

Source: article numbers reference the consolidated text at artificialintelligenceact.eu. Whether a given system is "high-risk" under the Act depends on its use; the questions above assume the obligations apply.

No surprises, no drip sequence

1

An email arrives

We send the download link to the address you give us — usually within a minute. Check spam if it's not there.

2

You download the PDF

20 questions, scored. Work through it with your team in 25 minutes. Every "no" is a gap to close before AI Act enforcement lands.

3

An optional follow-up

If you want to talk through your score, we're around. If you don't, you won't hear from us again. No drip sequence, no sales pressure.