EU AI Act · preliminary classification
What this system is, why, and what is still unknown
Answer what you know. The engine returns a preliminary classification with its reasoning and its gaps, and routes it to a person. It cannot settle the question — that is a human decision, recorded as a signature.
Determinative facts
- Is it an AI system within the Regulation?
- Is a prohibited practice under Article 5 suspected?
- Does the intended purpose fall in a high-risk area?
- Does a transparency duty under Article 50 apply?
- Do general-purpose model obligations apply?
- Does it act as a safety component?
- Are people affected by the output?
- Is personal data processed?
- Can a human meaningfully override the output?
- What does the output do?
- What role does the organisation play?
Preliminary result
Not yet determinable
Confidence in the facts Low — most determinative facts are missing
Why
- · No determinative signal has been recorded yet.
10 determinative facts not recorded
- Is it an AI system within the Regulation?
- Is a prohibited practice under Article 5 suspected?
- Does the intended purpose fall in a high-risk area?
- Transparency duty
- Actor role
- Does it act as a safety component?
- Are people affected by the output?
- What does the output do?
- Is personal data processed?
- Can a human meaningfully override the output?
A missing fact is not an absent risk. The classification cannot rise above the facts recorded, and none of these resolves in your favour by default.
Preliminary classification — human legal, regulatory or domain review required.
Owned for review by: AI_RISK_LEAD
What this engine will not do
- · Finalise a legal classification
- · Declare a system compliant
- · Approve a deployment
- · Treat a missing fact as an absent risk
- · Present a generated answer as verified fact