WorkflowEU AI Act

EU AI Act Choose a high-risk assessment route

Choose between provider-run internal control, notified-body assessment and a product-law procedure before release.

Start with the high-risk basis. Then check standards coverage, assessor requirements and the records needed to complete the route.

Author
Sorena AI
Published
May 9, 2026
Updated
Oct 21, 2026
Overview

Record why the system is high-risk before choosing a route. Article 43 treats Annex I Section A products, Annex III point 1 biometric systems and Annex III points 2-8 differently. Section B products follow the amended sector-law rules. Use the steps below to record the decision and finish the required market-entry checks.

Section 1

1. Record why the system is high-risk

Write down the high-risk classification first. Annex I product systems and Annex III use cases follow different procedures.

For Annex I, identify the section and sector legislation first. Section A includes New Legislative Framework legislation such as toys, radio equipment, pressure equipment, lifts, PPE, gas appliances, medical devices, and in vitro diagnostic medical devices.

Article 43 sends those systems through the procedure required by the sector law, with the AI Act Chapter III Section 2 requirements included in that assessment.

Regulation (EU) 2026/1744 moves machinery from Annex I Section A to Section B. For machinery and other Section B systems, only the AI Act provisions identified in amended Article 2(2) apply directly unless the relevant high-risk requirements have been integrated into the sector legislation.

Do not route a Section B system through Article 43 merely because it appears in Annex I.

For Annex III, record the exact point and use case. Annex III covers biometrics, critical infrastructure, education and vocational training, employment and workers' management, access to essential private and public services, law enforcement, migration and border control, and justice and democratic processes.

Route selection and application timing are separate questions.

Regulation (EU) 2026/1744 entered into force on 27 July 2026 and applies Chapter III Sections 1-3, except Article 6(5), from 2 December 2027 for systems classified under Article 6(2) and Annex III, and from 2 August 2028 for systems classified under Article 6(1) and Annex I.

Record the applicable route and date even when conformity work starts earlier.

  • Annex I route: the AI system is itself a regulated product or a safety component of a product covered by Annex I Union harmonisation legislation.
  • Annex III point 1 route: biometrics systems have special Article 43 triggers for notified body involvement.
  • Annex III points 2 to 8 route: internal control under Annex VI is the default Article 43 procedure unless the Commission later changes that routing by delegated act.
  • Evidence to capture: intended purpose, product or use-case category, applicable Annex item, provider identity, affected Member States, and the source used for the classification.
Section 2

2. Choose the Article 43 route

For Annex III point 1 biometric systems, provider-run internal control under Annex VI is available only when the provider fully applies harmonised standards or applicable covering the relevant requirements. Otherwise use Annex VII: a notified body assesses the quality management system and technical documentation.

Use Annex VII if relevant harmonised standards and are unavailable. Also use it when a standard is unused or only partly applied, applicable common specifications are unused, or a cited standard has a restriction affecting the relevant part.

For Annex III points 2 to 8, Article 43 points providers to the internal-control procedure in Annex VI. That still requires the provider to verify the quality management system, examine the technical documentation, and confirm that design, development, and post-market monitoring are consistent with the documentation.

For Annex I Section A product-law systems, follow the procedure required by the relevant product legislation.

If that legislation allows the manufacturer to opt out of third-party assessment because all relevant harmonised standards are applied, Article 43 allows that option only when harmonised standards or also cover all AI Act Chapter III Section 2 requirements.

If the same system is covered by Annex I Section A product legislation and also falls within an Annex III category, amended Article 43(3) selects the relevant product-law procedure. Record both classifications, then use the sector procedure rather than opening a separate Annex III route for the same system.

  • Route A - Annex I Section A product law: use the sector and include AI Act high-risk requirements in that assessment.
  • Route B - Annex III point 1 with full standards or : provider may choose Annex VI internal control or Annex VII assessment by a notified body.
  • Route C - Annex III point 1 without full standards or : use Annex VII assessment by a notified body.
  • Route D - Annex III points 2 to 8: use Annex VI internal control unless a later delegated act makes Annex VII applicable.
  • Special authority rule: when an Annex VII system is intended for law enforcement, immigration or asylum authorities, or Union institutions, bodies, offices, or agencies, Article 43 assigns the relevant market surveillance authority to act as the notified body, unless Article 75(1e) gives the AI Office responsibility for a system under its exclusive competence.
Section 3

3. Check standards coverage and assessor needs

EN 18286:2026 was cited on 14 October 2026 by Decision (EU) 2026/2274. Its presumption of conformity covers only the mapped requirements.

Annex ZA maps Article 17(1) and the first sentence of Article 11(1), with conditions: document every aspect in Article 17(1)(a)-(m), include every Annex IV element in the technical file, and maintain an Article 9-compliant risk-management system. Assess Article 17(2)-(4) separately because the standard does not cover them.

Applying EN 18286 alone does not cover all Chapter III Section 2 requirements or establish the Annex III point 1 conditions for Annex VI internal control. Check every relevant Section 2 requirement against the applied standards or ; retain Annex VII when Article 43 requires it.

The citation changes neither application dates nor the notified body's QMS and technical-documentation assessment duties. Follow the harmonised standards and quality-management guide for procedure and evidence mapping.

Standards coverage determines whether a point 1 biometric system can use internal control or needs Annex VII assessment. For each requirement, identify the harmonised standard used and whether its reference is published in the Official Journal of the European Union (OJEU).

Record full or partial application, any common specification used, and any alternative technical solution.

Harmonised standards and can create a presumption of conformity only to the extent that they cover the relevant AI Act requirements or obligations. If the provider does not comply with common specifications, Article 41 requires justification that the adopted technical solutions meet the requirements to at least an equivalent level.

When Annex VII applies, the evidence package must be ready for a notified body review of both the quality management system and the technical documentation.

Annex VII allows the notified body to request further evidence or tests, and in limited circumstances to access training, validation, and testing datasets or trained models where necessary for the assessment.

  • Standards record: standard or common specification, edition, OJEU decision and citation date, date checked, clause and Annex ZA mapping, covered requirement, full or partial application, restriction, system version, evidence link, owner, and residual requirement with its evidence and closure decision.
  • Notified body trigger record: Annex III point 1 trigger, standards gap, partial application, restricted standard, product-law requirement, or special public-authority route.
  • Annex VII submission record: quality management system documentation, Annex IV technical documentation, no-duplicate-application declaration, and the chosen notified body or relevant authority.
  • Alternative-solution record: technical solution, requirement covered, test evidence, risk-management link, and justification for equivalence where are not followed.
Section 4

4. Complete the declaration, marking and registration

Assign the remaining market-entry records before closing the route decision. Article 47 requires a written, machine-readable EU declaration of conformity for each high-risk system, signed physically or electronically. Keep it available to national authorities for 10 years after supplying or first using the system.

Article 48 requires CE marking for high-risk AI systems. Digital systems can use a digital CE marking if it is easily accessible through the interface, a machine-readable code, or another electronic means.

Where a notified body is responsible for the Article 43 , the CE marking must be followed by that body's identification number.

Registration depends on the high-risk basis and actor. Article 49 requires providers or authorised representatives to register Annex III high-risk systems, except point 2 critical infrastructure systems, in the EU database before placing them on the market or putting them into service.

Providers that classify an Annex III system as not high-risk under Article 6(3) must also register that conclusion. Public authorities and Union bodies deploying Annex III systems, except point 2, must register themselves and the use of the system. Article 49 routes point 2 critical infrastructure systems to national registration.

A new is required after a substantial modification. A planned change to a continuously learning system does not count as a substantial modification when the provider predetermined the change and its performance in the initial conformity assessment and included it in the technical documentation.

Store that predetermined-change evidence with the initial route decision.

Article 46 allows a market surveillance authority, on a duly justified request and for a limited period, to authorise a specific high-risk system before the conformity procedure is complete for exceptional public-security, life and health, environmental, or key industrial and infrastructure reasons.

For high-risk systems related to products covered by Annex I Section A, only the derogations in the applicable sector legislation apply. Treat this as an authority-controlled derogation, an exceptional authorisation: the required must still be completed without undue delay.

  • EU declaration record: system name and type, provider details, sole-responsibility statement, AI Act conformity statement, applicable data-protection statement, standards or , notified body and certificate details where applicable, place, date, signer, and signature.
  • CE marking record: marking location, digital access method where relevant, notified body identification number where applicable, and whether other Union law also requires CE marking.
  • Registration record: provider or deployer registration obligation, EU database or national registration path, Annex III exception check, certificate details where applicable, EU declaration copy, instructions for use, Member States, and status of the system.
  • Reassessment record: substantial modification, intended-purpose change, standards or common-specification change, notified body certificate supplement, and post-market monitoring evidence.
Primary sources

References and citations

eur-lex.europa.eu5 citations
Referenced sections
  • Primary legal source for Article 43 conformity assessment routes, Annex I and Annex III high-risk categories, Articles 40 and 41 standards and common specifications, Article 47 declaration, Article 48 CE marking, Article 49 registration, and Annex IV to VIII evidence fields.
  • Article 43 sets the conformity assessment route by high-risk category, while Annex I and Annex III define the product-law and use-case categories used by this selector.
  • Article 43 paragraphs 1 to 3 provide the internal-control, notified-body, and Annex I product-law routes used in the selector.
Show 2 more
  • Articles 40 and 41 explain presumption of conformity through harmonised standards and common specifications; Annex VII defines notified body review of the quality system and technical documentation.
  • Articles 47 to 49 and Annexes V and VIII support the declaration of conformity, CE marking, registration, and evidence fields listed here.
"For high-risk AI systems listed in point 1 of Annex III"
More quoted passages
"For high-risk AI systems covered by the Union harmonisation legislation listed in Section A of Annex I"
"the provider shall opt for one of the following conformity assessment procedures"
"shall be presumed to be in conformity with the requirements"
"The EU declaration of conformity shall identify the high-risk AI system"
digital-strategy.ec.europa.eu3 citations
Referenced sections
  • Commission overview supports the practical sequence from high-risk conformity assessment through possible notified body involvement, EU database registration, declaration of conformity, and CE marking.
  • Commission overview summarizes the provider sequence: conformity assessment, possible notified body involvement, EU database registration, declaration of conformity, and CE marking.
  • Commission overview confirms that, after conformity assessment and registration of standalone systems, a declaration of conformity is signed and the system bears CE marking before market placement.
"It needs to undergo the conformity assessment and comply with AI requirements."
More quoted passages
"For some systems a notified body is involved too."
"A declaration of conformity needs to be signed"
ai-act-service-desk.ec.europa.eu2 citations
Referenced sections
  • Official Article 49 page supports the registration obligations for Annex III high-risk systems, not-high-risk Article 6(3) conclusions, certain public deployers, restricted registration, and national registration for Annex III point 2 systems.
  • Official AI Act Service Desk article explains registration before placing Annex III high-risk systems on the market or putting them into service, including EU database and non-public registration cases.
"High-risk AI systems referred to in point 2 of Annex III shall be registered at national level."
More quoted passages
"Before placing on the market or putting into service a high-risk AI system"
cencenelec.eu2 citations
Referenced sections
  • Official standards-body publication announcement for the provider QMS standard.
  • Official publication announcement for the provider QMS standard; the later decision controls its citation status.
eur-lex.europa.eu2 citations
Referenced sections
  • Articles 1-2, recital 6 and the Annex establish the 14 October 2026 citation and its effect limited to covered requirements.
  • Articles 1-2, recital 6 and the Annex establish the 14 October 2026 citation and presumption limited to covered requirements.
digital-strategy.ec.europa.eu2 citations
Referenced sections
  • Commission FAQ supports the high-risk classification split between Annex I product-law systems and Annex III use cases, plus the need to repeat assessment after substantial modification.
  • Commission FAQ explains that high-risk classification depends on intended purpose and can arise from Annex I product legislation or Annex III use cases.
"The risk classification is based on the intended purpose of the AI system"
More quoted passages
"AI systems can classify as high-risk in two cases"
digital-strategy.ec.europa.eu2 citations
Referenced sections
  • Commission standardisation page supports the role of harmonised standards, OJEU references, voluntary use of standards, and presumption of compliance for providers applying referenced standards.
  • Commission standardisation page explains that harmonised standards are voluntary but, when referenced in the Official Journal, provide legal certainty and a presumption of compliance.
"harmonised standards referenced in the Official Journal of the EU provide legal certainty"
More quoted passages
"The application of standards remains voluntary."
eur-lex.europa.eu2 citations
Referenced sections
  • Binding amendment used for the revised high-risk application dates, Annex I Section A and Section B treatment, the transfer of machinery to Section B, and the AI Office's conformity-assessment responsibility under Article 75(1e).
  • Binding amendment supporting the 2 December 2027 and 2 August 2028 dates, the Section A and Section B distinction, and the transfer of machinery to Section B.
Show all related guides (34)

EU AI Act serious incident reporting steps

Check serious-incident harm, preserve evidence and choose the reporting deadline and recipient for high-risk AI systems and systemic-risk GPAI models.

EU AI Act timeline and phased deadlines

See which AI Act rules apply now and what comes next. Check high-risk dates, older-model transitions and the difference between legal and standards dates.

EU AI Act vs ISO/IEC 42001

Compare AI Act legal duties with ISO/IEC 42001 management-system requirements. See which records can be reused and what certification does and does not cover.

EU AI Act vs NIST AI RMF

Compare AI Act legal duties with the voluntary NIST AI RMF. Map risk records to specific requirements and identify legal steps that need separate evidence.

EU AI Act: Check whether the Act applies

Use five checks to establish AI Act scope, exclusions, your role, risk categories and evidence for the decision.

EU AI Act: Classify a tool before approval

Classify an AI tool or model before purchase or release. Check EU scope, prohibited uses, high-risk rules, transparency and GPAI duties.

EU AI Act: Compliance checklist

Check each AI tool, decide which duties apply, and give every action an owner and evidence link.

EU AI Act: Conformity assessment

Some high-risk systems use a provider-run assessment. Others need an authorised independent assessor or a product-law assessment.

EU AI Act: Does it apply to you?

Check whether the AI Act applies to your system or model, identify your provider or deployer role, and keep evidence for the decision.

EU AI Act: Duties for general-purpose AI models

Check GPAI provider duties for documentation, integrator information, copyright and training summaries, plus extra requirements for systemic-risk models.

EU AI Act: High-risk requirements

High-risk providers must meet system requirements before supply. Deployers must use and monitor the system under the provider's instructions.

EU AI Act: High-risk system checklist

Check the system's risk controls, data, documentation, logs, instructions, human oversight and testing before release.

EU AI Act: Is your use case high-risk?

Find high-risk AI examples in recruitment, banking, education, healthcare, public services and products. Check Article 6 conditions and exceptions.

EU AI Act: Plan your compliance work

Build a compliance plan for each AI tool: your role, the duties that apply, the evidence needed and the person responsible.

EU AI Act: Standards and quality management

Use EN 18286:2026 to organise provider quality-management procedures and keep evidence that they work for each high-risk system.

EU AI Act: Technical file and evidence templates

Build the technical file and provider records for a high-risk AI system, with owners and evidence links for each field.

EU AI Act: When must you disclose AI?

Check Article 50 chatbot notices, synthetic-output marking, biometric notices, deepfake disclosures and public-interest text exceptions.

EU AI Act: Which AI uses are banned?

Check Article 5 prohibited AI practices, the narrow exceptions, evidence to retain and the new content prohibitions from 2 December 2026.

Are you the provider or the deployer? | EU AI Act

Check provider, deployer, importer and distributor roles, when changes create provider duties, and how model and system responsibilities differ.

Do you need a notified body? | EU AI Act

Find the assessment route for high-risk AI, when a notified body is required, and which declaration, CE-marking and registration records you need.

Does this tool count as AI? | EU AI Act

Check whether software counts as AI, how models differ from systems, when embedded AI is high-risk, and which scope and role evidence to keep.

How do you monitor AI after release? | EU AI Act

Understand high-risk AI monitoring under Article 72: the provider plan, deployer feedback, logs, corrective action, incidents and change reviews.

Is this industry use high-risk? | EU AI Act

See which industry AI uses fall under Annex III, when Article 6(3) can apply, and what providers must document and register.

What belongs in the technical file? | EU AI Act

See what Article 11 and Annex IV require in high-risk AI technical documentation, from system identity and testing to monitoring and retention.

What must general-purpose model providers do? | EU AI Act

Check general-purpose model duties, systemic-risk classification, open-source conditions, copyright, training summaries, incidents and application dates.

When must you disclose AI use? | EU AI Act

Check AI interaction notices, technical content marking, biometric notices, deepfake and text disclosures, exceptions, accessibility and deadlines.

When must you report a serious incident? | EU AI Act

Check serious-incident outcomes, reporting roles, 15-, 10- and two-day deadlines, authority recipients, corrective action and the separate model route.

Who needs a rights impact assessment? | EU AI Act

Check who needs an Article 27 fundamental rights impact assessment, what it must cover, and how privacy assessments, notification and registration fit.