Skip to content
AIRAS Cloud

Free readiness check

How ready is your organisation for the EU AI Act?

Ten questions, three minutes, no sign-up. Answer from what you could evidence today and you will get an honest maturity band plus the exact themes to fix first.

Runs in your browser Nothing stored or sent Ten EU AI Act themes

Score your current position

Choose the option that reflects what you could show an auditor this week, not what your policy promises.

  1. 01 · Inventory

    Do you hold a complete, current inventory of every AI system in use?

    Including embedded vendor features, copilots and models inside SaaS tools.

    Do you hold a complete, current inventory of every AI system in use?
  2. 02 · Shadow AI

    Can you detect AI adopted by teams without central approval?

    Unregistered tools are the single most common source of unmanaged exposure.

    Can you detect AI adopted by teams without central approval?
  3. 03 · Regulatory role

    Have you determined your role for each AI system under the EU AI Act?

    Provider, deployer, importer or distributor determines which obligations apply.

    Have you determined your role for each AI system under the EU AI Act?
  4. 04 · Prohibited practices

    Have you screened your AI use against the Article 5 prohibitions?

    Prohibited practices carry the highest penalties and were the first to apply.

    Have you screened your AI use against the Article 5 prohibitions?
  5. 05 · Risk classification

    Do you know which of your systems are likely to be high-risk?

    Annex III use cases and safety components drive the heaviest obligations.

    Do you know which of your systems are likely to be high-risk?
  6. 06 · Transparency

    Do you meet Article 50 disclosure duties where they apply?

    Chatbots, synthetic content, emotion recognition and deepfake disclosure.

    Do you meet Article 50 disclosure duties where they apply?
  7. 07 · Human oversight

    Is human oversight defined, assigned and evidenced per system?

    Named accountable people, escalation thresholds and a record of interventions.

    Is human oversight defined, assigned and evidenced per system?
  8. 08 · Monitoring

    Do you monitor AI systems and agents after go-live?

    Post-market monitoring, drift, incidents and material-change detection.

    Do you monitor AI systems and agents after go-live?
  9. 09 · Evidence

    Could you produce an inspection-ready evidence pack within five working days?

    Assessment records, decisions, approvals, oversight logs and audit trail.

    Could you produce an inspection-ready evidence pack within five working days?
  10. 10 · Accountability

    Is a named executive accountable for AI governance outcomes?

    Segregation of duties between business owners, reviewers and approvers.

    Is a named executive accountable for AI governance outcomes?

From score to evidence

What closing each gap actually looks like

01

Find every system

Document discovery and structured intake surface the AI already in use, including embedded vendor features and unapproved tools, and place each one under a named owner.

Discovery and intake

02

Assess it deterministically

Role determination, applicability, prohibited-practice screening, transparency duties and high-risk pre-classification run through a deterministic engine with a full explanation trace.

How assessment works

03

Prove it on demand

Approvals, oversight records, monitoring signals and material-change detection accumulate into an append-only record that exports as an inspection-ready evidence pack.

Evidence and trust

Questions

About the readiness check

Last reviewed 2 August 2026.

Turn your score into a defensible record.

Bring your result to a 30-minute working session. We will map your weakest themes onto the AIRAS Cloud workflow and show exactly what the evidence pack would contain for your own AI systems.

No pricing commitment. No confidential information required.