Comparison

Article 50 versus Annex III: what applies now and what gets heavier later

Article 50 is a transparency duty that stands apart from the risk level and that became applicable earlier on the timeline. Annex III determines when a system is high-risk and thereby opens the heavier requirement set, at a later point on that same timeline.

The official source remains authoritative. This is general interpretation and not legal advice about your situation.

The fields side by side

Every row comes from the objects themselves. Where a field is empty, it says so; we do not fill in an assumption where the source is silent.

Article 50 versus Annex III: what applies now and what gets heavier later
Field	Article 50, any risk level	Annex III, high-risk
Status	Applicable	Upcoming
Applies from	2 August 2026	2 December 2027
Who carries the duty	Deployer, Provider of an AI system	Provider of an AI system
Who is affected	Not recorded	Deployer
Who supervises	Not recorded	Not recorded
When this applies	An AI system is intended to interact directly with natural persons. The system generates or manipulates synthetic audio, image, video or text, or the use concerns a specifically listed deployer scenario.	The intended purpose falls within a use case listed in Annex III. Classification follows Article 6(2).
Exceptions	The direct-interaction disclosure is not required where this is obvious to a reasonably well-informed, observant and circumspect person, considering circumstances and context. Only Article 50(2) has a transition until 2 December 2026 for systems placed on the market before 2 August 2026.	A listed Annex III system may fall outside high-risk under the strict conditions in Article 6(3), except where it profiles. The assessment and registration must be documented.
First actions	Implement the applicable disclosure, marking or label	Classify the use case and document the outcome
Evidence that belongs with it	Transparency implementation record	Article 6 and Annex III classification record
Control	Pre-release transparency check	Reclassification on purpose or context change
Version and review status	v1.0.0, placed against the official source	v1.0.0, placed against the official source

Status

Article 50, any risk level: Applicable
Annex III, high-risk: Upcoming

Applies from

Article 50, any risk level: 2 August 2026
Annex III, high-risk: 2 December 2027

Who carries the duty

Article 50, any risk level: Deployer, Provider of an AI system
Annex III, high-risk: Provider of an AI system

Who is affected

Article 50, any risk level: Not recorded
Annex III, high-risk: Deployer

Who supervises

Article 50, any risk level: Not recorded
Annex III, high-risk: Not recorded

When this applies

Article 50, any risk level: An AI system is intended to interact directly with natural persons. The system generates or manipulates synthetic audio, image, video or text, or the use concerns a specifically listed deployer scenario.
Annex III, high-risk: The intended purpose falls within a use case listed in Annex III. Classification follows Article 6(2).

Exceptions

Article 50, any risk level: The direct-interaction disclosure is not required where this is obvious to a reasonably well-informed, observant and circumspect person, considering circumstances and context. Only Article 50(2) has a transition until 2 December 2026 for systems placed on the market before 2 August 2026.
Annex III, high-risk: A listed Annex III system may fall outside high-risk under the strict conditions in Article 6(3), except where it profiles. The assessment and registration must be documented.

First actions

Article 50, any risk level: Implement the applicable disclosure, marking or label
Annex III, high-risk: Classify the use case and document the outcome

Evidence that belongs with it

Article 50, any risk level: Transparency implementation record
Annex III, high-risk: Article 6 and Annex III classification record

Control

Article 50, any risk level: Pre-release transparency check
Annex III, high-risk: Reclassification on purpose or context change

Version and review status

Article 50, any risk level: v1.0.0, placed against the official source
Annex III, high-risk: v1.0.0, placed against the official source

Article 50: transparency Annex III: high-risk AI

Official sources and locators

Per side, the provisions the statements above rest on.

Article 50, any risk level

EU Artificial Intelligence Act 2024/1689
Locator: Article 50(1)-(5) and Article 113
Open official source
Guidelines on Article 50
Locator: Final guidelines, scope by Article 50 paragraph
Open official source
Guidelines on Article 50
Locator: Implementation guidance for providers and deployers
Open official source

Annex III, high-risk

Digital Omnibus on AI 2026/1744
Locator: Amended Article 113, Article 6(2) and Annex III application date
Open official source
EU Artificial Intelligence Act 2024/1689
Locator: Article 6 and Annex III
Open official source
EU Artificial Intelligence Act 2024/1689
Locator: Article 6(2)-(4), Article 49 and Annex III
Open official source

Referring to these two objects

Each side has its own stable identifier, version and hash. Take them separately; you cite a comparison by citing the two objects.

Referring to this object

Citation block

Copy this reference into your advice, article or file. The identifier, the version and the hash keep the statement findable later, even once the dataset has moved on.

Reference

Praxikon, "Article 50: transparency",
praxikon:eu:ai-act:obligation:article-50-transparency@1.0.0,
dataset praxikon:sys:registry:dataset:ai-act-implementation-graph 2.1.0 (schema 1.4.0),
effective_at 2026-08-02T00:00:00.000Z, known_at 2026-08-08T00:00:00.000Z,
sha256 239fbac0e4728dc239352b2f88b199c3cc098082ff72e2972b4b5e8a2b121406,
https://www.praxikon.com/en/verplichtingen/article-50-transparency
(https://www.praxikon.com/api/v1/obligations?id=praxikon%3Aeu%3Aai-act%3Aobligation%3Aarticle-50-transparency&effective_at=2026-08-02&known_at=2026-08-08&lang=en, accessed 2026-08-24)

Short form

praxikon:eu:ai-act:obligation:article-50-transparency@1.0.0 (sha256 239fbac0)

BibTeX

@misc{praxikon-eu-ai-act-obligation-article-50-transparency-1-0-0,
  author       = {{Praxikon}},
  title        = {Article 50: transparency},
  year         = {2026},
  version      = {1.0.0},
  number       = {praxikon:eu:ai-act:obligation:article-50-transparency},
  howpublished = {AI Act Change \& Evidence Graph, dataset 2.1.0, schema 1.4.0},
  note         = {effective_at 2026-08-02T00:00:00.000Z; known_at 2026-08-08T00:00:00.000Z; sha256 239fbac0e4728dc239352b2f88b199c3cc098082ff72e2972b4b5e8a2b121406},
  url          = {https://www.praxikon.com/en/verplichtingen/article-50-transparency},
  urldate      = {2026-08-24},
  language     = {en}
}

CSL JSON

[
  {
    "id": "praxikon:eu:ai-act:obligation:article-50-transparency@1.0.0",
    "type": "dataset",
    "title": "Article 50: transparency",
    "container-title": "AI Act Change & Evidence Graph",
    "publisher": "Praxikon",
    "version": "1.0.0",
    "number": "praxikon:eu:ai-act:obligation:article-50-transparency",
    "URL": "https://www.praxikon.com/en/verplichtingen/article-50-transparency",
    "language": "en",
    "issued": {
      "date-parts": [
        [
          2026,
          8,
          8
        ]
      ]
    },
    "accessed": {
      "date-parts": [
        [
          2026,
          8,
          24
        ]
      ]
    },
    "note": "dataset praxikon:sys:registry:dataset:ai-act-implementation-graph 2.1.0; schema 1.4.0; effective_at 2026-08-02T00:00:00.000Z; known_at 2026-08-08T00:00:00.000Z; sha256 239fbac0e4728dc239352b2f88b199c3cc098082ff72e2972b4b5e8a2b121406; retrieved_from https://www.praxikon.com/api/v1/obligations?id=praxikon%3Aeu%3Aai-act%3Aobligation%3Aarticle-50-transparency&effective_at=2026-08-02&known_at=2026-08-08&lang=en; licence https://www.praxikon.com/nl/legal/terms"
  }
]

How to verify a reference later is set out in the methodology. Terms

Referring to this object

Citation block

Copy this reference into your advice, article or file. The identifier, the version and the hash keep the statement findable later, even once the dataset has moved on.

Reference

Praxikon, "Annex III: high-risk AI",
praxikon:eu:ai-act:obligation:annex-iii-high-risk@1.0.0,
dataset praxikon:sys:registry:dataset:ai-act-implementation-graph 2.1.0 (schema 1.4.0),
effective_at 2026-07-27T00:00:00.000Z, known_at 2026-08-08T00:00:00.000Z,
sha256 c0afd1789ed393ba3f9ce04205bd74b4831ff0fd58146108fdd8dd08d2f4f6c9,
https://www.praxikon.com/en/verplichtingen/annex-iii-high-risk
(https://www.praxikon.com/api/v1/obligations?id=praxikon%3Aeu%3Aai-act%3Aobligation%3Aannex-iii-high-risk&effective_at=2026-07-27&known_at=2026-08-08&lang=en, accessed 2026-08-24)

Short form

praxikon:eu:ai-act:obligation:annex-iii-high-risk@1.0.0 (sha256 c0afd178)

BibTeX

@misc{praxikon-eu-ai-act-obligation-annex-iii-high-risk-1-0-0,
  author       = {{Praxikon}},
  title        = {Annex III: high-risk AI},
  year         = {2026},
  version      = {1.0.0},
  number       = {praxikon:eu:ai-act:obligation:annex-iii-high-risk},
  howpublished = {AI Act Change \& Evidence Graph, dataset 2.1.0, schema 1.4.0},
  note         = {effective_at 2026-07-27T00:00:00.000Z; known_at 2026-08-08T00:00:00.000Z; sha256 c0afd1789ed393ba3f9ce04205bd74b4831ff0fd58146108fdd8dd08d2f4f6c9},
  url          = {https://www.praxikon.com/en/verplichtingen/annex-iii-high-risk},
  urldate      = {2026-08-24},
  language     = {en}
}

CSL JSON

[
  {
    "id": "praxikon:eu:ai-act:obligation:annex-iii-high-risk@1.0.0",
    "type": "dataset",
    "title": "Annex III: high-risk AI",
    "container-title": "AI Act Change & Evidence Graph",
    "publisher": "Praxikon",
    "version": "1.0.0",
    "number": "praxikon:eu:ai-act:obligation:annex-iii-high-risk",
    "URL": "https://www.praxikon.com/en/verplichtingen/annex-iii-high-risk",
    "language": "en",
    "issued": {
      "date-parts": [
        [
          2026,
          8,
          8
        ]
      ]
    },
    "accessed": {
      "date-parts": [
        [
          2026,
          8,
          24
        ]
      ]
    },
    "note": "dataset praxikon:sys:registry:dataset:ai-act-implementation-graph 2.1.0; schema 1.4.0; effective_at 2026-07-27T00:00:00.000Z; known_at 2026-08-08T00:00:00.000Z; sha256 c0afd1789ed393ba3f9ce04205bd74b4831ff0fd58146108fdd8dd08d2f4f6c9; retrieved_from https://www.praxikon.com/api/v1/obligations?id=praxikon%3Aeu%3Aai-act%3Aobligation%3Aannex-iii-high-risk&effective_at=2026-07-27&known_at=2026-08-08&lang=en; licence https://www.praxikon.com/nl/legal/terms"
  }
]

How to verify a reference later is set out in the methodology. Terms

Execution

Decide per use case which transparency measure you take

Which of the two transparency duties applies depends on your role and on who is on the other side. Praxikon does not carry that out; Embed AI maps that per AI use case and sets the measure and the record alongside it.

See the Embed AI approach

Unsure about your role itself rather than the duties that come with it? Work out whether you are a provider or a deployer

The fields side by side

Official sources and locators

Article 50, any risk level

Annex III, high-risk

Referring to these two objects

Citation block

Citation block

Decide per use case which transparency measure you take

Other comparisons