Independent and not affiliated with the FDA, MHRA, ISPE, PDA, or any agency. Get the appgoutham@madhadi.com
madhadi.comData Integrity & GxP Quality
Browse all topics → Articles Templates & Procedures Learning paths GlossaryScenariosToolsRegulatory ReferencesLearning PathsTopics About Start here
Risk Assessment Plug-and-play starting point AI & Automation

Risk Assessment: EU AI Act Risk-Tier Classification and Non-High-Risk Justification

A plug-and-play classification record for one AI system under the EU AI Act: the prohibited-use screen, the two high-risk routes, the Article 6(3) non-high-risk exception with its Article 6(4) documented justification, role determination, and the resulting obligations, with a filled specimen and the articles it satisfies.

Document type: Risk Assessment

Read and copy the template below into your own quality system. It is a generic starting point for your own internal use, provided as is, with no warranty; see the Terms and License. Adopting it does not by itself create compliance.

This is a ready-to-use classification record for a single AI system under the EU AI Act (Regulation (EU) 2024/1689). Replace every <<FILL: ...>> placeholder with your own specifics, set your document numbers and dates, and route it through your normal document control, review, and approval. A worked filled specimen follows the template. This is general guidance to adapt, not legal or regulatory advice; confirm the Act’s text, articles, and dates against the current Official Journal version before relying on a classification.

Document control header

FieldEntry
Record titleEU AI Act Classification: <<FILL: system name>>
Document number<<FILL: CLASS-AI-ID>>
Version<<FILL>>
Date<<FILL>>
Linked register entry<<FILL: register System ID>>
Owner<<FILL: AI Governance Lead>>

1. System description and intended purpose

FieldEntry
System name<<FILL>>
What it does<<FILL: function, inputs, outputs, the decision it informs>>
Degree of autonomy<<FILL: advisory only / influences a human decision / acts without human review>>
Where output is used<<FILL: include whether used in the EU>>
Develop or buy<<FILL: built in-house / commercial / embedded>>

2. Step 1: Prohibited-use screen (Article 5)

Screen first, because this is the highest-penalty band.

Prohibited category (Article 5)Applies?Note
Social scoring<<FILL: No / Flag>><<FILL>>
Untargeted facial-image scraping<<FILL>><<FILL>>
Manipulative or exploitative systems<<FILL>><<FILL>>
Emotion inference in the workplace (narrow medical/safety exception)<<FILL>><<FILL: check HR tools>>
Prohibited biometric categorisation<<FILL>><<FILL: check security tools>>
Screen resultPass / Flag<<FILL>>

If any category is flagged, stop and escalate to Legal; do not deploy pending resolution.

3. Step 2: High-risk determination (Article 6)

Route 1, Annex I (Article 6(1)): AI inside a regulated product.

QuestionAnswer
Is the AI a safety component of, or itself, a product under listed EU product-safety law (e.g. MDR (EU) 2017/745 or IVDR (EU) 2017/746)?<<FILL: Yes / No>>
Does that product require third-party (notified-body) conformity assessment? (broadly MDR Class IIa+ or IVDR Class B+)<<FILL: Yes / No>>
Annex I high-risk?<<FILL: Yes if both Yes>>

Route 2, Annex III (Article 6(2)): listed standalone uses (e.g. employment and worker management, access to essential services, biometric identification, critical infrastructure).

QuestionAnswer
Does the system fall in an Annex III category?<<FILL: Yes / No, which>>
Annex III high-risk (subject to the exception below)?<<FILL>>

4. Step 2a: Article 6(3) non-high-risk exception (only if Annex III applied)

A provider may treat an Annex III system as not high-risk if it does not pose a significant risk of harm. This section is the Article 6(4) documented justification and must be completed before the system goes on the market.

Exception condition (Article 6(3))Met?Justification
Performs a narrow procedural task<<FILL>><<FILL>>
Only improves the result of completed human activity<<FILL>><<FILL>>
Detects patterns / deviations without replacing or influencing the human assessment<<FILL>><<FILL>>
Performs a preparatory task to an assessment<<FILL>><<FILL>>
Does the system profile natural persons? (if yes, always high-risk, exception unavailable)<<FILL: No required to use exception>><<FILL>>
ConclusionNot high-risk / High-risk<<FILL: the reasoned basis>>

5. Step 3 and 4: Limited and minimal risk (if not high-risk)

QuestionAnswer
Does it interact with people or generate/manipulate content?<<FILL: Yes -> limited risk (transparency duties) / No>>
If not prohibited, high, or limited: minimal risk<<FILL>>

6. Role determination

QuestionAnswer
Do you develop it or have it built and put it on the market under your name?<<FILL: Yes -> Provider>>
Do you use it under your own authority in a professional context?<<FILL: Yes -> Deployer>>
Did you substantially modify or retrain a bought system, changing its purpose or behaviour?<<FILL: Yes -> you may become the Provider of a new system>>
Role<<FILL: Provider / Deployer / Both>>

7. Resulting obligations and owners

Obligation area (scaled to tier and role)Applies?Internal owner
Risk-management system across lifecycle<<FILL>><<FILL>>
Data governance and data-quality for training/validation/test data<<FILL>>Data Science
Technical documentation (kept current)<<FILL>><<FILL>>
Automatic logging / record-keeping<<FILL>><<FILL>>
Human oversight design<<FILL>><<FILL>>
Accuracy, robustness, cybersecurity<<FILL>><<FILL>>
Quality management system<<FILL>>Quality
Conformity assessment and CE marking (high-risk)<<FILL>>Regulatory Affairs
Registration and post-market monitoring<<FILL>><<FILL>>
Transparency (limited risk: disclose AI / mark content)<<FILL>><<FILL>>
GDPR lawful basis and DPIA<<FILL>>Legal / Privacy

8. Conclusion

FieldEntry
Final classification<<FILL: Prohibited / High-risk (Annex I or III) / Limited / Minimal>>
Role<<FILL>>
Applicable deadline (verify current)<<FILL>>
Re-classification triggersRetraining, repurposing, new intended use, change of provider, Act amendment

9. References

Regulation (EU) 2024/1689: Article 5; Article 6(1) and Annex I; Article 6(2) and Annex III; Article 6(3) and 6(4); transparency obligations (Article 50). EU MDR (EU) 2017/745 and IVDR (EU) 2017/746 for the device-constituent route. GDPR (EU) 2016/679 for the personal-data overlap. Confirm articles and dates against the current Official Journal text; the Digital Omnibus package may alter dates and details.

10. Approvals

RoleNameSignatureDate
Author (AI Governance)<<FILL>>
Reviewer (Legal/Regulatory)<<FILL>>
Approver (Quality)<<FILL>>

Filled specimen

Classification of an example operational quality model (illustrative).

FieldEntry
SystemDeviation-triage model that ranks open deviations by suggested investigation priority
AutonomyAdvisory: ranks deviations already opened by humans; humans decide and act
Article 5 screenPass
Annex I routeNo (not a regulated product)
Annex III routeNot in an Annex III category (not employment, essential services, or biometric)
Article 6(3)Not reached (Annex III did not apply); even if it had, the model only improves the result of completed human activity and does not profile persons
Limited risk?No (does not interact with the public or generate content)
Final classificationMinimal risk
RoleProvider and Deployer (built in-house for own use)
ObligationsNo mandatory AI-Act obligations; voluntary good practice applies. Documented here so the conclusion survives a question.

Contrast with the connected dosing device feature: assessed on its own the device constituent is MDR Class IIa and needs a notified body, so the AI is high-risk by the Annex I route, the role is provider, and the full conformity regime applies, assessed inside the MDR procedure (see matrix-ai-act-mdr-ivdr-crosswalk).

Common findings this record prevents

  • A “minimal-risk” or “not-high-risk” conclusion with no written reasoning.
  • Using the Article 6(3) exception without the Article 6(4) documented assessment made before market.
  • Misreading the role: treating yourself as a deployer when you built or retrained the model.
  • Skipping the prohibited-use screen on HR and security tools.
  • Treating MDR/IVDR conformity as if it covered the AI-Act requirements.

How to adapt this record

  1. Complete one record per AI system and link it to the inventory register row.
  2. Run the steps in order: prohibited screen, high-risk routes, exception, transparency.
  3. Record the Article 6(4) justification as controlled data when you rely on the exception.
  4. Assign each resulting obligation to a named internal owner.
  5. Confirm the articles and deadlines against the current Official Journal text.
Use madhadi.com as an app Full screen, works offline, one tap from your home screen.