{"meta":{"register_id":"praxikon:sys:registry:register:corrections","dataset_id":"praxikon:sys:registry:dataset:ai-act-implementation-graph","dataset_version":"2.8.0","schema_version":"1.5.0","lang":"en","register_started_at":"2026-08-08T00:00:00.000Z","count":52,"complete":true,"filters":{"id":null,"entity_id":null}},"data":[{"id":"praxikon:sys:registry:correction:2026-11-20-article-4a-replaces-article-10-5","corrected_at":"2026-11-20T00:00:00.000Z","severity":"material","previous_statement":"Second, Article 10(5) contains an exception permitting processing of special category data where strictly necessary to detect and correct bias in high-risk systems. [...] It applies only to high-risk systems [...]. Record whether you rely on Article 10(5), for which high-risk system [...]. Where the AI Act does open space, such as Article 10(5) [...]. For special categories of personal data the strict exception of Article 10(5) applies (bias detection and correction under conditions).","corrected_statement":"Second, Article 4a contains an exception permitting processing of special category data where strictly necessary to detect and correct bias. Paragraph 1, formerly Article 10(5), applies to providers of high-risk systems [...]. Paragraph 2 is new and gives the same room, without creating an obligation, to providers and deployers of other AI systems and models and to deployers of high-risk systems [...]. Record whether you rely on Article 4a, on paragraph 1 or on paragraph 2, in which role and for which system [...]. Also include those reasons in your records of processing activities, as Article 4a(1), point (f), requires. [...] Where the AI Act does open space, such as Article 4a, whose paragraph 1 was formerly Article 10(5) [...]. For special categories of personal data the strict exception of Article 4a(1), formerly Article 10(5), applies (bias detection and correction under conditions).","reason":"Article 1, point (9)(b), of Regulation (EU) 2026/1744 deleted Article 10(5) of the AI Act; Article 1, point (6), inserted Article 4a. Since 27 July 2026 the exception for processing special categories of personal data to detect and correct bias is in Article 4a. Paragraph 1 applies to providers of high-risk AI systems; paragraph 2 also applies, under the same conditions and safeguards, to providers and deployers of other AI systems and models and to deployers of high-risk AI systems. The three objects carry known date 8 August 2026, after the amendment: from the start they wrongly named the deleted paragraph as applicable law, and the sentence that the exception applies only to high-risk systems was no longer correct. Found on 20 November 2026 while clearing references to deleted provisions; the new text was independently reviewed the same day.","affected_entity_ids":["praxikon:eu:ai-act:definition:definitie-bijzondere-categorieen-persoonsgegevens","praxikon:eu:ai-act:definition:definitie-persoonsgegevens","praxikon:eu:ai-act:obligation:article-10-data-governance"],"affected_paths":["/en/verkenner/definition/definitie-bijzondere-categorieen-persoonsgegevens","/en/verkenner/definition/definitie-persoonsgegevens","/en/verkenner/obligation/article-10-data-governance","/nl/verkenner/definition/definitie-bijzondere-categorieen-persoonsgegevens","/nl/verkenner/definition/definitie-persoonsgegevens","/nl/verkenner/obligation/article-10-data-governance"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.8.0"},{"id":"praxikon:sys:registry:correction:2026-11-19-annex-i-section-b-applicable-articles","corrected_at":"2026-11-19T00:00:00.000Z","severity":"material","previous_statement":"For Section B only Article 6(1), Articles 102 to 109 and Article 112 apply. [...] The draft guidelines refer to Article 2(2) AI Act for this.","corrected_statement":"For Section B, since Regulation (EU) 2026/1744, only Article 6(1), Article 60a and Articles 102 to 112 apply; Articles 57, 58 and 59 apply only in so far as the requirements for high-risk systems have been integrated in that harmonisation legislation. [...] The draft guidelines refer to Article 2(2) AI Act for this, in the text as it stood before Regulation (EU) 2026/1744. [...] The same regulation moved machinery: point 1 of Section A was deleted and Regulation (EU) 2023/1230 was added to Section B as point 21.","reason":"Article 1, point (2)(a), of Regulation (EU) 2026/1744 replaces Article 2(2). Since 27 July 2026 only Article 6(1), Article 60a and Articles 102 to 112 apply to high-risk AI systems related to products under the harmonisation legislation of Annex I, Section B, and Articles 57, 58 and 59 only in so far as the requirements have been integrated in that harmonisation legislation. The object was published on 8 August 2026 and its summary still gave the text of 12 July 2024. The rendering of the draft guidelines of 19 May 2026 (point 60) was accurate and stays, with the note that it predates the amendment. Found during the legal review of 19 November 2026.","affected_entity_ids":["praxikon:eu:ai-act:guidance:guidance-high-risk-annex-i-section-a-versus-section-b"],"affected_paths":["/en/verkenner/guidance/guidance-high-risk-annex-i-section-a-versus-section-b","/nl/verkenner/guidance/guidance-high-risk-annex-i-section-a-versus-section-b"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.7.0"},{"id":"praxikon:sys:registry:correction:2026-11-19-article-41-1-conditions-not-settled","corrected_at":"2026-11-19T00:00:00.000Z","severity":"material","previous_statement":"That is precisely the state described by the conditions of Article 41(1), point (a)(ii), and point (b), and so the common specification route remains practically relevant.","corrected_statement":"That fits the condition of Article 41(1), point (a)(ii) (the requested standards were not delivered within the deadline of the request), and the first part of point (b) for the requirements of Section 2 for which no reference has been published. Whether point (b) is met is not settled by that: it also requires that no such reference is expected to be published within a reasonable period. That judgement lies with the Commission, which under paragraph 2 first informs the committee referred to in Article 22 of Regulation (EU) No 1025/2012 that it considers the conditions fulfilled. The common specification route is therefore a real possibility, not a given.","reason":"Article 41(1), point (b), of the consolidated text of 27 July 2026 requires that no reference to harmonised standards covering the requirements of Section 2 has been published \"and no such reference is expected to be published within a reasonable period\". Paragraph 2 has the Commission first inform the committee referred to in Article 22 of Regulation (EU) No 1025/2012 that it considers the conditions fulfilled. The note read the present state as \"precisely\" the state of point (a)(ii) and point (b), and skipped that second part of point (b). On point (a)(ii): request C(2025) 3871 asked CEN and Cenelec to draft the standards by 31 August 2025 (Article 1(1) of that decision). Whether the conditions of paragraph 1 are fulfilled has not been published; the note now says it is not settled. The summary of the obligation now also names the conditions of Article 41(1) instead of only the absence of a standard. Hence material: a condition for a power was presented as fulfilled.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-40-42-standards-and-specifications"],"affected_paths":["/en/verkenner/obligation/article-40-42-standards-and-specifications","/nl/verkenner/obligation/article-40-42-standards-and-specifications"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/2026-07-27/nld","dataset_version":"2.6.10"},{"id":"praxikon:sys:registry:correction:2026-11-19-article-60-annex-i-section-a-and-article-60a","corrected_at":"2026-11-19T00:00:00.000Z","severity":"material","previous_statement":"Applies where you are a provider or prospective provider of an Annex III high-risk AI system and you want to test it in real world conditions outside an AI regulatory sandbox [...] Article 60(1) provides that testing of high-risk AI systems in real world conditions outside AI regulatory sandboxes may be conducted by providers or prospective providers of high-risk AI systems listed in Annex III [...]","corrected_statement":"Applies where you are a provider or prospective provider of a high-risk AI system listed in Annex III, or of a high-risk AI system covered by the Union harmonisation legislation listed in Annex I, Section A [...] Article 60 does not apply to high-risk AI systems related to products under the harmonisation legislation of Annex I, Section B: Article 2(2) applies only Article 6(1), Article 60a and Articles 102 to 112 to those systems. [...]","reason":"Article 1, point (24), of Regulation (EU) 2026/1744 replaces the first subparagraph of Article 60(1) and Article 60(2). Since 27 July 2026 providers or prospective providers of high-risk AI systems covered by the Union harmonisation legislation listed in Section A of Annex I may also test outside a sandbox under Article 60. The third subparagraph of paragraph 1 was not amended. Article 1, point (25), inserts Article 60a for products under Annex I, Section B, and Article 2(2) applies Article 60a, not Article 60, to those systems. The obligation named Annex III only. Hence material: the scope in the condition, the exception and the official fact moves. The official fact now also cites Article 1, points (24) and (25), and recital 26 of 2026/1744.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-60-real-world-testing"],"affected_paths":["/en/verkenner/obligation/article-60-real-world-testing","/nl/verkenner/obligation/article-60-real-world-testing"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.11"},{"id":"praxikon:sys:registry:correction:2026-11-19-decision-2026-2274-recital-1-article-40","corrected_at":"2026-11-19T00:00:00.000Z","severity":"editorial","previous_statement":"The decision does not name any articles. [...] The decision names no articles.","corrected_statement":"The decision refers to Article 40 in Recital 1, but does not say which requirements of the AI Act the standard covers.","reason":"Recital 1 of Implementing Decision (EU) 2026/2274 opens with \"In accordance with Article 40 of Regulation (EU) 2024/1689\". The decision therefore does name an Article; what it does not do is say which requirements of the AI Act EN 18286:2026 covers. The point made, that the decision does not fix the coverage of the standard, stays the same. In the post on Annex III and Articles 7 and 8, recital 3 is now also marked as a recital of the decision. Hence editorial.","affected_entity_ids":[],"affected_paths":["/en/posts/amending-annex-iii-compliance-requirements-high-risk-ai-article-7-8-ai-act","/en/posts/quality-management-system-high-risk-ai-article-17-ai-act","/nl/posts/kwaliteitsbeheer-hoog-risico-ai-artikel-17-ai-act","/nl/posts/wijziging-bijlage-iii-naleving-eisen-hoog-risico-ai-artikel-7-8-ai-act"],"source_url":"https://eur-lex.europa.eu/eli/dec_impl/2026/2274/oj/nld","dataset_version":"2.6.10"},{"id":"praxikon:sys:registry:correction:2026-11-19-en-18286-presumption-article-40-1","corrected_at":"2026-11-19T00:00:00.000Z","severity":"material","previous_statement":"From that date, compliance with the standard confers under Article 40(1) of Regulation (EU) 2024/1689 a presumption of conformity with the requirements the standard covers. [...] Since 14 October 2026 EN 18286 has what it lacked before: a legal presumption. [...] so applying that standard does confer a presumption of conformity for the requirements it covers.","corrected_statement":"Article 40(1) ties the presumption to the requirements of Chapter III, Section 2, and to the obligations of Chapter V, Sections 2 and 3, to the extent the standard covers them; Article 17 (quality management system) sits in Chapter III, Section 3. Regulation (EU) 2026/1744 added a subparagraph to Article 40(2) [...] (Article 1, point (17)); it did not amend Article 40(1). [...] Whether the presumption also extends to Article 17, the subject of the standard, is not settled.","reason":"Article 40(1) of the consolidated text of 27 July 2026 confers the presumption of conformity with \"the requirements set out in Section 2 of this Chapter\" or the obligations of Chapter V, Sections 2 and 3, \"to the extent that those standards cover those requirements or obligations\". EN 18286:2026 concerns the quality management system of Article 17, and that Article sits in Chapter III, Section 3. Article 1, point (17), of Regulation (EU) 2026/1744 added a subparagraph to Article 40(2) in which the Commission requests standardisation deliverables to facilitate \"the joint compliance and presumption of conformity with the requirements or obligations set out in Chapter III, Sections 2 and 3\", but left paragraph 1 unchanged. Implementing Decision (EU) 2026/2274 refers to Article 40 in recital 1 and does not say which requirements the standard covers. Five objects stated without reservation that the standard confers a presumption under Article 40(1), which could lead a provider to believe that applying EN 18286 presumes compliance with Article 17. The new versions state the presumption to the extent the standard covers requirements that Article 40(1) names, and the coverage of Article 17 as an open point; the note on Articles 40 to 42 now calls the coverage statement accompanying the standard unconfirmed, as the standard object already did. Hence material: the reach of a legal effect moves. Beside standardisation request M/613 the new versions also give the official reference, Implementing Decision C(2025) 3871 of 23 June 2025 (recital 2 of 2026/2274; the Commission's eNorm register: \"C(2025)3871 - Standardisation request M/613\").","affected_entity_ids":["praxikon:eu:ai-act:change:2026-10-14-en-18286-harmonised","praxikon:eu:ai-act:definition:definitie-geharmoniseerde-norm","praxikon:eu:ai-act:obligation:article-40-42-standards-and-specifications","praxikon:eu:ai-act:standard:standard-en-18286-quality-management-system","praxikon:eu:ai-act:standard:standard-en-iso-iec-42001-ai-management-system"],"affected_paths":["/en/verkenner/change/2026-10-14-en-18286-harmonised","/en/verkenner/definition/definitie-geharmoniseerde-norm","/en/verkenner/obligation/article-40-42-standards-and-specifications","/en/verkenner/standard/standard-en-18286-quality-management-system","/en/verkenner/standard/standard-en-iso-iec-42001-ai-management-system","/nl/verkenner/change/2026-10-14-en-18286-harmonised","/nl/verkenner/definition/definitie-geharmoniseerde-norm","/nl/verkenner/obligation/article-40-42-standards-and-specifications","/nl/verkenner/standard/standard-en-18286-quality-management-system","/nl/verkenner/standard/standard-en-iso-iec-42001-ai-management-system"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/2026-07-27/nld","dataset_version":"2.6.10"},{"id":"praxikon:sys:registry:correction:2026-11-19-gpai-guidelines-adopted-text-point-numbers","corrected_at":"2026-11-19T00:00:00.000Z","severity":"material","previous_statement":"The Commission guidelines of 18 July 2025 (C(2025) 5045 final) [...] state in point (61) [...] Point (63) sets the indicative criterion [...] Point (64) [...] Point (70) states that where a downstream actor modifies a model classified as having systemic risk [...] point (71) [...]","corrected_statement":"The Commission guidelines [...], formally adopted on 19 November 2025 (C(2025) 7719 final), state in point (58) [...] Point (60) sets the indicative criterion [...] Point (61) [...] Point (67) states that where a downstream actor modifies a model with systemic risk [...] point (68) [...]","reason":"The objects cited the text the Commission approved in content on 18 July 2025, C(2025) 5045 final. The Commission formally adopted the guidelines on 19 November 2025 as C(2025) 7719 final. In the adopted text the paragraph numbers of Section 3.2 moved three places: Section 3.2 runs from (57) to (64), Sections 3.2.1 and 3.2.2 from (65) to (68). The objects were published on or after 8 August 2026, so after the adoption. A reader who looked up the cited point in the adopted text arrived at a different paragraph; hence material. There is one difference in content: adopted point (67) speaks of a general-purpose AI model with systemic risk, not of a model classified as such. Eight objects receive an additional version: the guidance object on the downstream modifier, the Article 52 obligation, five examples (the locator only) and the change of 18 July 2025 (a sentence on the formal adoption). Points (57) to (68) and (78) to (81) were compared against the adopted text on 19 November 2026.","affected_entity_ids":["praxikon:eu:ai-act:change:2025-07-18-gpai-guidelines","praxikon:eu:ai-act:example:example-finetunen-boven-de-compute-drempel","praxikon:eu:ai-act:example:example-lichte-finetuning-geen-modelaanbieder","praxikon:eu:ai-act:example:example-model-met-systeemrisico-aanpassen","praxikon:eu:ai-act:example:example-opensource-in-naam-maar-niet-in-de-zin-van-de-verordening","praxikon:eu:ai-act:example:example-opensource-uitzondering-reikwijdte","praxikon:eu:ai-act:guidance:guidance-gpai-downstream-modifier-becomes-provider","praxikon:eu:ai-act:obligation:article-52-systemic-risk-classification"],"affected_paths":["/en/verkenner/change/2025-07-18-gpai-guidelines","/en/verkenner/guidance/guidance-gpai-downstream-modifier-becomes-provider","/en/verkenner/obligation/article-52-systemic-risk-classification","/nl/verkenner/change/2025-07-18-gpai-guidelines","/nl/verkenner/guidance/guidance-gpai-downstream-modifier-becomes-provider","/nl/verkenner/obligation/article-52-systemic-risk-classification"],"source_url":"https://digital-strategy.ec.europa.eu/en/policies/guidelines-gpai-providers","dataset_version":"2.7.0"},{"id":"praxikon:sys:registry:correction:2026-11-18-article-11-no-deployer-right-to-the-file","corrected_at":"2026-11-18T00:00:00.000Z","severity":"material","previous_statement":"For buying organisations, Annex IV is the checklist of what you must be able to request contractually from your supplier; without that file you cannot meet your own Article 26 duties. [...] Include the right to access the technical documentation in AI procurement and supplier contracts now.","corrected_statement":"The Regulation gives the deployer no right to the Annex IV technical documentation: Article 11(1) has it drawn up so that national competent authorities and notified bodies can assess compliance. The deployer works with the instructions for use that accompany the system under Article 13, and Article 26(1) requires the deployer to use the system in accordance with those instructions. Access to parts of Annex IV is something the deployer can only negotiate by contract. [...] If, as a buying organisation, you need more than the Article 13 instructions for use, negotiate access to the parts of the technical documentation you need in AI procurement and supplier contracts. That is a contractual choice; the Regulation does not require it.","reason":"No provision of Regulation (EU) 2024/1689 gives the deployer a right to the technical documentation. Article 11(1) has it drawn up so that national competent authorities and notified bodies have the information necessary to assess compliance. What the deployer receives are the instructions for use: Article 13(2): \"High-risk AI systems shall be accompanied by instructions for use\", and Article 26(1) requires deployers to use the system \"in accordance with the instructions for use accompanying the systems\". The statement that a buying organisation cannot meet its Article 26 duties without the Annex IV file, and the recommended action that spoke of \"the right to access\", could lead a deployer to believe that the law grants it that file or that it cannot comply without it. Hence material, although no condition, date, addressee or status moves. The recommended action stays as a contractual choice and now says that the Regulation does not require it; both statements now also cite Article 13 and Article 26(1).","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-11-technical-documentation"],"affected_paths":["/en/verkenner/obligation/article-11-technical-documentation","/nl/verkenner/obligation/article-11-technical-documentation"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.6.8"},{"id":"praxikon:sys:registry:correction:2026-11-18-article-11-simplified-form","corrected_at":"2026-11-18T00:00:00.000Z","severity":"material","previous_statement":"Small providers (SMEs) may provide the documentation in the simplified form established by the Commission.","corrected_statement":"SMEs, including start-ups, and small mid-cap enterprises may provide the Annex IV elements in a simplified manner. If they opt to do so, they use the simplified technical documentation form that the Commission is to establish under Article 11(1). The documentation duty itself still applies.","reason":"Article 11(1), second subparagraph, as amended by Regulation (EU) 2026/1744, reads since 27 July 2026: \"SMEs, including start-ups, and SMCs, may provide the elements of the technical documentation specified in Annex IV in a simplified manner. To that end, the Commission shall establish a simplified technical documentation form targeted at the needs of SMEs, including start-ups, and SMCs.\" The exception did not name small mid-cap enterprises, spoke of \"small providers\" and of \"the simplified form established by the Commission\", as if that form already exists. A small mid-cap enterprise could conclude that the simplification is not open to it, and an SME could look for a form the Regulation only announces. Hence material, although no condition, date or status moves. The new text says that the Commission is to establish the form, without saying that it has been published, and that the documentation duty itself still applies.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-11-technical-documentation"],"affected_paths":["/en/verkenner/obligation/article-11-technical-documentation","/nl/verkenner/obligation/article-11-technical-documentation"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.8"},{"id":"praxikon:sys:registry:correction:2026-11-18-article-49-annex-viii-section-b","corrected_at":"2026-11-18T00:00:00.000Z","severity":"editorial","previous_statement":"It is paid for in visibility: your name, your system and the Article 6(3) condition you rely on end up in a publicly searchable register, exactly where you thought you would stay out of sight. The mistake that follows is predictable and expensive.","corrected_statement":"It is paid for in visibility: your name, your system and the Article 6(3) condition you rely on end up in a publicly searchable register, exactly where you thought you would stay out of sight. What that entry must contain is set out in Annex VIII, Section B, and since 27 July 2026 that is less. Article 1, point (42), of Regulation (EU) 2026/1744 deletes points 7 and 9 there: the short summary of the grounds on which the AI system is considered to be not high-risk in application of the procedure under Article 6(3), and any Member States in which the system has been placed on the market, put into service or made available in the Union. The condition or conditions under Article 6(3) that you rely on remain in the registration as point 6. [...] Your reasoning therefore no longer goes into the database, but it must be ready. Article 49(4), point (b), still names point 9 of Section B for the secure section; the amending Regulation did not adjust that reference, so it points to a deleted point. The mistake that follows is predictable and expensive.","reason":"Article 1, point (42), of Regulation (EU) 2026/1744 reads: \"in Annex VIII, section B, points 7 and 9 are deleted\". Since 27 July 2026 the registration under Article 49(2) therefore no longer asks for a short summary of the grounds on which the system is not high-risk in application of Article 6(3), nor for the Member States. Recital 22 calls that a simplification and maintains that the provider must still document its assessment and that national competent authorities should be able to request it. The statement on paragraph 2 did not mention that amendment. What it said was not wrong, but a reader of the statement did not learn that the entry has become shorter. Hence editorial: no date, status, addressee or condition moves. The statement now also cites Article 1, point (42), and recital 22 of 2026/1744.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-49-registration"],"affected_paths":["/en/verkenner/obligation/article-49-registration","/nl/verkenner/obligation/article-49-registration"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.9"},{"id":"praxikon:sys:registry:correction:2026-11-17-article-63-only-two-places","corrected_at":"2026-11-17T00:00:00.000Z","severity":"editorial","previous_statement":"Together they are the only two places in this Regulation where the size of an undertaking legally makes a difference, and neither touches the substantive requirements for high-risk AI systems.","corrected_statement":"Neither touches the substantive requirements for high-risk AI systems.","reason":"Articles 62 and 63 are not the only places where the size of an undertaking legally makes a difference. Article 11(1), second subparagraph, lets SMEs, including start-ups, and small mid-cap enterprises provide the Annex IV technical documentation in a simplified manner, and Article 99(6) makes the lower of the two fine ceilings decisive for SMEs, including start-ups. The sentence is removed; the finding that Articles 62 and 63 do not touch the substantive requirements for high-risk AI systems stays. Editorial, because it is an editorial statement and the scope of Article 63 does not move.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-63-sme-derogations"],"affected_paths":["/en/verkenner/obligation/article-63-sme-derogations","/nl/verkenner/obligation/article-63-sme-derogations"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.7"},{"id":"praxikon:sys:registry:correction:2026-11-17-article-63-sme-scope","corrected_at":"2026-11-17T00:00:00.000Z","severity":"material","previous_statement":"Test whether you are a microenterprise with no partner or linked enterprises [...] File on microenterprise status [...] A microenterprise providing a high-risk system therefore does very nearly everything a large company does [...] A microenterprise invoking Article 63 [...] are you a microenterprise within the meaning of Recommendation 2003/361/EC","corrected_statement":"Test whether you are an SME, including a start-up, with no partner or linked enterprises [...] File on SME status [...] An SME providing a high-risk system therefore does very nearly everything a large company does [...] An SME invoking Article 63 [...] are you an SME, including a start-up, within the meaning of Recommendation 2003/361/EC","reason":"Article 1, point (26), of Regulation (EU) 2026/1744 replaced Article 63(1); recital 28 says that the option should be extended to all SMEs, including start-ups. Since 27 July 2026 paragraph 1 reads: \"SMEs, including start-ups, may comply with certain elements of the quality management system required by Article 17 in a simplified manner, provided that they do not have partner enterprises or linked enterprises within the meaning of Recommendation 2003/361/EC.\" The obligation already rendered that in its summary, condition and official fact, but the action, the evidence object and three editorial sentences of the obligation still named only the microenterprise. A standalone small or medium-sized enterprise following those sentences could wrongly conclude that it cannot rely on Article 63. Hence material, although no condition, date, addressee or status moves.","affected_entity_ids":["praxikon:eu:ai-act:action:article-63-scope-simplified-quality-management","praxikon:eu:ai-act:evidence:article-63-sme-eligibility-record","praxikon:eu:ai-act:obligation:article-63-sme-derogations"],"affected_paths":["/en/verkenner/action/article-63-scope-simplified-quality-management","/en/verkenner/evidence/article-63-sme-eligibility-record","/en/verkenner/obligation/article-63-sme-derogations","/nl/verkenner/action/article-63-scope-simplified-quality-management","/nl/verkenner/evidence/article-63-sme-eligibility-record","/nl/verkenner/obligation/article-63-sme-derogations"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.7"},{"id":"praxikon:sys:registry:correction:2026-11-17-glossary-ai-office-competence","corrected_at":"2026-11-17T00:00:00.000Z","severity":"material","previous_statement":"The European AI Office is the body within the European Commission responsible for implementing the EU AI Act. The AI Office develops guidelines, coordinates with national supervisors, and oversees GPAI models and prohibited practices.","corrected_statement":"The AI Office is a function of the European Commission, established in 2024 within the Directorate-General CNECT; references in the AI Act to the AI Office are to be understood as references to the Commission (Article 3(47)). The Commission develops Union expertise through the AI Office (Article 64). It enforces the obligations of providers of general-purpose AI models and facilitates codes of practice. It already had powers over AI systems where the model and the system come from the same provider; the Digital Omnibus extended and specified that competence: the AI Office is now exclusively competent for AI systems that a provider builds on its own model or a model from the same group, and for AI systems in very large online platforms and search engines, with exceptions (Article 75(1)). All other AI systems, including for the prohibited practices, and deployers that are neither the provider nor part of the same undertaking are supervised by national market surveillance authorities; Union institutions by the European Data Protection Supervisor (Article 74(9)).","reason":"The old definition said without qualification that the AI Office oversees prohibited practices. That holds only for the AI systems in Article 75(1) AI Act, as amended by Regulation (EU) 2026/1744. For all other AI systems the national market surveillance authorities enforce Article 5, and the AI Office's exclusive competence applies to deployers only when they are also the provider or form part of the same undertaking (Article 75(1), second subparagraph; recital (31) of Regulation (EU) 2026/1744). The English glossary page also pointed to an article that does not exist; it now points to the existing enforcement article and the new article on the AI Office.","affected_entity_ids":[],"affected_paths":["/en/glossary/ai-office","/nl/glossary/ai-office"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.6"},{"id":"praxikon:sys:registry:correction:2026-11-17-glossary-social-scoring-syri-fines","corrected_at":"2026-11-17T00:00:00.000Z","severity":"material","previous_statement":"The Dutch SyRI system, which a court in The Hague struck down in 2020 for violating the ECHR, shows how close such systems can come to prohibited social scoring. Violations fall into the heaviest fine category: up to 35 million euros or 7% of global annual turnover, imposed by the supervisory authority.","corrected_statement":"In 2020 the District Court of The Hague declared the legislation for the Dutch SyRI system, a risk model for benefit, allowance and tax fraud, non-binding because it conflicted with Article 8(2) ECHR (ECLI:NL:RBDHA:2020:865). That judgment concerned the right to private life, not the AI Act, but it raises the same questions: which data a risk model uses and whether the consequences are proportionate. Violations fall into the heaviest fine category: up to 35 million euros or 7% of worldwide annual turnover, whichever is higher (Article 99(3)). Member States lay down the rules on penalties, and depending on the legal system a court may impose the fine (Article 99(1) and (9)).","reason":"The District Court of The Hague did not ban SyRI: it declared section 65 of the SUWI Act and chapter 5a of the SUWI Decree non-binding because they conflicted with Article 8(2) ECHR (ECLI:NL:RBDHA:2020:865, legal ground 7.2). The judgment did not apply the AI Act, which did not exist yet, and says nothing about how close SyRI came to the social scoring ban. The English glossary page also said the fine is imposed by the supervisory authority; under Article 99(1) and (9) AI Act the Member States lay down the penalty rules and, depending on the legal system, a court may impose the fine. Both glossary pages now also link to the new article on social scoring.","affected_entity_ids":[],"affected_paths":["/en/glossary/social-scoring","/nl/glossary/sociale-scoring"],"source_url":"https://uitspraken.rechtspraak.nl/details?id=ECLI:NL:RBDHA:2020:865","dataset_version":"2.6.6"},{"id":"praxikon:sys:registry:correction:2026-11-13-article-4a-article-2-7-reservation","corrected_at":"2026-11-13T00:00:00.000Z","severity":"material","previous_statement":"Until 27 July 2026 that paragraph left the GDPR unaffected without reservation [...] The Union legislature therefore carved the reservation out for Article 4a precisely [...] The previous version of that paragraph carried no such reservation for Article 4a.","corrected_statement":"Already in the original text of Regulation (EU) 2024/1689 that paragraph made a reservation for the basis in its former place: the Regulation left the GDPR unaffected \"without prejudice to Article 10(5) and Article 59 of this Regulation\". [...] That the Union legislature carried the reservation over to Article 4a fits the reading that Article 4a is itself the Union law measure [...], but the reservation itself already existed for Article 10(5).","reason":"Article 2(7) of Regulation (EU) 2024/1689 read, in the text of 12 July 2024: \"This Regulation shall not affect Regulation (EU) 2016/679 or (EU) 2018/1725, or Directive 2002/58/EC or (EU) 2016/680, without prejudice to Article 10(5) and Article 59 of this Regulation.\" Regulation (EU) 2026/1744, Article 1, point 2(b), therefore only moved the reservation from Article 10(5) to Article 4a, together with the basis. The note used the reservation as a new argument for reading Article 4a as the Union law measure itself; it no longer counts as a new argument, and both the chosen reading and the alternative reading stay. The practical sentence on older documents now names the replaced wording \"without prejudice to Article 10(5)\".","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-4a-bias-testing-legal-basis"],"affected_paths":["/en/verkenner/obligation/article-4a-bias-testing-legal-basis","/nl/verkenner/obligation/article-4a-bias-testing-legal-basis"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.6.6"},{"id":"praxikon:sys:registry:correction:2026-11-13-article-4a-official-dutch-text","corrected_at":"2026-11-13T00:00:00.000Z","severity":"editorial","previous_statement":"The Dutch rendering of Article 4a(1) and (2) used its own words, with \"vertekening\" for bias and, under point (d), \"niet aan andere partijen doorgegeven, overgedragen of anderszins toegankelijk gemaakt\".","corrected_statement":"The Dutch rendering gives the official Dutch text, with \"vooringenomenheid\" and, under point (d), \"worden niet verzonden, doorgegeven of anderszins geraadpleegd door andere partijen\". The English text did not change.","reason":"The Dutch rendering of Article 4a(1) and (2) was presented as verbatim but used its own words, including \"vertekening\" where the official Dutch text has \"vooringenomenheid\", and a different wording under point (d). The two official facts now give the text of Article 4a as inserted by Article 1, point 6, of Regulation (EU) 2026/1744 (data/ai-act/review/sources/reg-eu-2026-1744-nl.txt). The summary, conditions, exception, the fact on recital 9 and the editorial statement follow the same words. The English text was already verbatim. The substance of the object does not change.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-4a-bias-testing-legal-basis"],"affected_paths":["/en/verkenner/obligation/article-4a-bias-testing-legal-basis","/nl/verkenner/obligation/article-4a-bias-testing-legal-basis"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.6"},{"id":"praxikon:sys:registry:correction:2026-11-12-article-6-paragraphs-change-three-inserted","corrected_at":"2026-11-12T00:00:00.000Z","severity":"editorial","previous_statement":"Article 1, point (8), of Regulation (EU) 2026/1744 inserts the following paragraphs into Article 6: 1a. [...] 1b. [...]","corrected_statement":"Article 1, point (8), of Regulation (EU) 2026/1744 inserts three paragraphs, 1a to 1c, into Article 6. The first two read: 1a. [...] 1b. [...]","reason":"Article 1, point (8), inserts three paragraphs into Article 6: 1a, 1b and 1c (data/ai-act/review/sources/reg-eu-2026-1744-nl.txt and -en.txt). The fact named \"the following paragraphs\" and gave two, so it read as if there were only two. The quoted text of 1a and 1b was right and stays; the summary already described paragraph 1c. The substance of the object does not change.","affected_entity_ids":["praxikon:eu:ai-act:change:2026-07-27-safety-component-narrowed"],"affected_paths":["/en/verkenner/change/2026-07-27-safety-component-narrowed","/nl/verkenner/change/2026-07-27-safety-component-narrowed"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.5"},{"id":"praxikon:sys:registry:correction:2026-11-12-article-6-paragraphs-change-title","corrected_at":"2026-11-12T00:00:00.000Z","severity":"editorial","previous_statement":"Article 6 gains paragraphs 1a to 1c on safety components (locator of the official fact: Inserted Article 6(1a)-(1c)).","corrected_statement":"Article 6 gains paragraphs 1a to 1c: safety component and third-party conformity assessment (locator of the official fact: Inserted Article 6(1a) and (1b)).","reason":"Regulation (EU) 2026/1744, Article 1, point (8), inserts three paragraphs into Article 6. Only paragraphs 1a and 1b concern the notion of safety component; paragraph 1c provides that a product required to undergo a third-party conformity assessment solely due to risks other than risks to health and safety shall not be considered as fulfilling the condition in paragraph 1, point (b). The title said that all three paragraphs concern safety components. The summary already described paragraph 1c correctly, so the substance of the object is unchanged. The official fact quotes only paragraphs 1a and 1b, and the locator now points to those two paragraphs. Read in data/ai-act/review/sources/reg-eu-2026-1744-nl.txt and -en.txt.","affected_entity_ids":["praxikon:eu:ai-act:change:2026-07-27-safety-component-narrowed"],"affected_paths":["/en/verkenner/change/2026-07-27-safety-component-narrowed","/nl/verkenner/change/2026-07-27-safety-component-narrowed"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.4"},{"id":"praxikon:sys:registry:correction:2026-11-12-article-6-route-article-3-14-quotation","corrected_at":"2026-11-12T00:00:00.000Z","severity":"editorial","previous_statement":"The English quotation of the amended Article 3, point (14), ended: \"to prevent or mitigate risks to the health and safety of persons or property\".","corrected_statement":"The English quotation now ends word for word as the official text does: \"to prevent or mitigate risks to health and safety of persons or property\".","reason":"An official fact quotes word for word. The English text of Article 3, point (14), as amended by Regulation (EU) 2026/1744, Article 1, point (4)(a), has no article before health (data/ai-act/review/sources/reg-eu-2026-1744-en.txt and reg-eu-2024-1689-consolidated-20260727-en.txt). The substance does not change. The Dutch quotation matches both Dutch source files word for word and is unchanged.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-6-1bis-1quater-route"],"affected_paths":["/en/verkenner/obligation/article-6-1bis-1quater-route","/nl/verkenner/obligation/article-6-1bis-1quater-route"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.4"},{"id":"praxikon:sys:registry:correction:2026-11-12-article-6-route-inserted-paragraphs-locator","corrected_at":"2026-11-12T00:00:00.000Z","severity":"editorial","previous_statement":"Locator of the inserted paragraphs: Inserted Article 6(1a)-(1c).","corrected_statement":"Locator: Article 1, point (8), inserting Article 6(1a) to (1c).","reason":"The locator named the target provision but not where the amending Regulation inserts it: Article 1, point (8), of Regulation (EU) 2026/1744. The three citations in the module all concern the three paragraphs together. Review finding R6B-2. Only the reference changes.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-6-1bis-1quater-route"],"affected_paths":["/en/verkenner/obligation/article-6-1bis-1quater-route","/nl/verkenner/obligation/article-6-1bis-1quater-route"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.5"},{"id":"praxikon:sys:registry:correction:2026-11-12-article-6-route-note-definition-object","corrected_at":"2026-11-12T00:00:00.000Z","severity":"editorial","previous_statement":"The published definition object on Article 3, point (14), still carries the unamended text and the reading that a component need not have an identifiable safety function.","corrected_statement":"Since the correction of 12 November 2026 the published definition object on Article 3, point (14), gives the amended text, but it still carries the reading that a component need not have an identifiable safety function.","reason":"The note described the state of another object in the dataset. Since dataset 2.6.3 that object (version 1.0.1) gives the amended text of Article 3, point (14); the editorial reading on it is unchanged. Version 1.0.1 of the module, of the same day, still carried the old sentence. The legal reading in the note does not change.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-6-1bis-1quater-route"],"affected_paths":["/en/verkenner/obligation/article-6-1bis-1quater-route","/nl/verkenner/obligation/article-6-1bis-1quater-route"],"source_url":null,"dataset_version":"2.6.5"},{"id":"praxikon:sys:registry:correction:2026-11-12-efficiency-diaeresis-in-quotations","corrected_at":"2026-11-12T00:00:00.000Z","severity":"editorial","previous_statement":"The Dutch quotations of Article 6(1a) and Article 51(3) read \"efficientie\".","corrected_statement":"The quotations spell \"efficiëntie\", as the source does.","reason":"An official fact quotes word for word. The Dutch text of Article 6(1a) (data/ai-act/review/sources/reg-eu-2026-1744-nl.txt) and of Article 51(3) (reg-eu-2024-1689-consolidated-20260727-nl.txt) spells \"efficiëntie\". Only the spelling changes.","affected_entity_ids":["praxikon:eu:ai-act:change:2026-07-27-safety-component-narrowed","praxikon:eu:ai-act:obligation:article-52-systemic-risk-classification","praxikon:eu:ai-act:obligation:article-6-1bis-1quater-route"],"affected_paths":["/en/verkenner/change/2026-07-27-safety-component-narrowed","/en/verkenner/obligation/article-52-systemic-risk-classification","/en/verkenner/obligation/article-6-1bis-1quater-route","/nl/verkenner/change/2026-07-27-safety-component-narrowed","/nl/verkenner/obligation/article-52-systemic-risk-classification","/nl/verkenner/obligation/article-6-1bis-1quater-route"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.5"},{"id":"praxikon:sys:registry:correction:2026-11-12-safety-component-definition-2026-1744","corrected_at":"2026-11-12T00:00:00.000Z","severity":"material","previous_statement":"Article 3(14) defines a safety component as a component of a product or of an AI system which fulfils a safety function for that product or AI system, or the failure or malfunctioning of which endangers the health and safety of persons or property.","corrected_statement":"The same definition, with the clause Regulation (EU) 2026/1744 added to it: for the purposes of this definition, a component fulfils a safety function where its intended purpose is to prevent or mitigate risks to health and safety of persons or property.","reason":"Regulation (EU) 2026/1744, Article 1, point (4)(a), amends Article 3, point (14), and sets out when a component fulfils a safety function. The amendment has operated since its entry into force on 27 July 2026, so the published version with a reference date of 8 August 2026 was incomplete on its own reference date. The clause helps decide whether an AI system falls under Article 6(1) or Annex III, point 2. Read in the consolidated text of 27 July 2026 and in the amending Regulation; the provision module for Article 6(1a) to (1c) already carried the amended text. Found while re-reading the review file (finding R6B-1).","affected_entity_ids":["praxikon:eu:ai-act:definition:definitie-veiligheidscomponent"],"affected_paths":["/en/verkenner/definition/definitie-veiligheidscomponent","/nl/verkenner/definition/definitie-veiligheidscomponent"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.6.3"},{"id":"praxikon:sys:registry:correction:2026-11-11-article-49-template-applicable","corrected_at":"2026-11-11T00:00:00.000Z","severity":"material","previous_statement":"The template with the full text of Article 49 carried the status not yet applicable (upcoming).","corrected_statement":"The template carries the status applicable, as the obligation, action, evidence and control of Article 49 do: Article 49 has applied since 2 August 2026.","reason":"Decision D1 of 6 September 2026 chose the text date: Article 49 sits in Section 5 of Chapter III, which appears in none of the points of the third paragraph of Article 113, so the general date of 2 August 2026 applies. The four other objects of the module moved to applicable; the template was not on the list of affected objects and stayed behind. Found while re-reading the review file on 11 November 2026 (finding B49-4).","affected_entity_ids":["praxikon:eu:ai-act:template:article-49-legal-text"],"affected_paths":["/en/verkenner/template/article-49-legal-text","/nl/verkenner/template/article-49-legal-text"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.6.1"},{"id":"praxikon:sys:registry:correction:2026-11-11-article-86-article-111-paragraph-1","corrected_at":"2026-11-11T00:00:00.000Z","severity":"material","previous_statement":"For a system already in use, Article 111(2) applies: the Regulation then applies only once its design changes significantly, and for systems of public authorities by 2 August 2030 at the latest.","corrected_statement":"The same sentence, adding that paragraph 2 excludes the systems of paragraph 1: AI systems which are components of the large-scale IT systems listed in Annex X and were placed on the market or put into service before 2 August 2027 must, under Article 111(1), be brought into compliance by 31 December 2030.","reason":"Article 111(2) applies to operators of high-risk AI systems \"other than the systems referred to in paragraph 1 of this Article\". Those systems have their own deadline in paragraph 1. The note left that exception out, although in practice it can reach systems in point 7 of Annex III. Read in the consolidated text of 27 July 2026; Regulation (EU) 2026/1744 replaces only paragraph 2 and leaves paragraph 1 unchanged. Found while re-reading the review file on 11 November 2026 (finding R86-4).","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-86-right-to-explanation"],"affected_paths":["/en/verkenner/obligation/article-86-right-to-explanation","/nl/verkenner/obligation/article-86-right-to-explanation"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.6.2"},{"id":"praxikon:sys:registry:correction:2026-11-11-implementing-decision-2026-2274-date","corrected_at":"2026-11-11T00:00:00.000Z","severity":"editorial","previous_statement":"Of the twelve standard objects only EN 18286 has that publication, since Implementing Decision (EU) 2026/2274 of 14 October 2026.","corrected_statement":"Of the twelve standard objects only EN 18286 has that publication, since Implementing Decision (EU) 2026/2274 of 12 October 2026, published on 14 October 2026.","reason":"The decision is dated 12 October 2026 and was published in the Official Journal on 14 October 2026 (title, Article 2 and closing formula of the archived text). The presumption of conformity runs from publication, so the substance was right; only the date of the decision was wrong. Found while re-reading the review file on 11 November 2026 (finding B4042-4).","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-40-42-standards-and-specifications"],"affected_paths":["/en/verkenner/obligation/article-40-42-standards-and-specifications","/nl/verkenner/obligation/article-40-42-standards-and-specifications"],"source_url":"https://eur-lex.europa.eu/eli/dec_impl/2026/2274/oj","dataset_version":"2.6.1"},{"id":"praxikon:sys:registry:correction:2026-11-06-ai-agent-disclosure-on-whose-behalf","corrected_at":"2026-11-06T00:00:00.000Z","severity":"material","previous_statement":"AI agents must disclose both their AI nature and on whose behalf they act; the obligation is twofold: not only 'I am AI', but also 'I act on behalf of this party'.","corrected_statement":"The binding duty is the disclosure in Article 50(1): people are informed that they are interacting with an AI system, unless that is obvious. The non-binding Commission guidelines (C(2026) 5054, point 31) additionally ask that an agent also states on whose behalf it acts.","reason":"Article 50(1) of Regulation (EU) 2024/1689 does not mention on whose behalf an AI system acts. That comes from point 31 of the guidelines, and point 5 says they are not binding. The label, summary and interpretation presented it as a duty under the law. Read in EUR-Lex and in the guidelines on 6 November 2026.","affected_entity_ids":["praxikon:eu:ai-act:guidance:guidance-article-50-ai-agent-self-disclosure"],"affected_paths":["/en/posts/ai-agents-gdpr-article-22-automated-decisions","/en/verkenner/guidance/guidance-article-50-ai-agent-self-disclosure","/nl/posts/ai-agents-artikel-22-avg-geautomatiseerde-besluiten","/nl/verkenner/guidance/guidance-article-50-ai-agent-self-disclosure"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.5.1"},{"id":"praxikon:sys:registry:correction:2026-10-27-article-57-sandbox-deadline","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"The published text provides that the sandbox shall be operational by 2 August 2026; the consolidated text has not been verified.","corrected_statement":"Since Regulation (EU) 2026/1744 the sandbox must be operational by 2 August 2027; that is how Article 57(1) reads in the consolidated text of 27 July 2026.","reason":"The consolidated text has been in the repository since early September, but the object kept the old date as a fact and a source label still said 2 August 2026.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-57-regulatory-sandboxes"],"affected_paths":["/en/richtsnoeren/vormvoorstel-nederlandse-regulatory-sandbox-onder-de-ai-verordening","/en/verkenner/obligation/article-57-regulatory-sandboxes","/nl/richtsnoeren/vormvoorstel-nederlandse-regulatory-sandbox-onder-de-ai-verordening","/nl/verkenner/obligation/article-57-regulatory-sandboxes"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-calendar-next-deadline","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"The next AI Act deadline is 2 August 2027 (countdown widget); 2 August 2026 is listed as a fixed, future date.","corrected_statement":"The next date is 2 December 2026: end of the transition period for machine-readable marking (Article 50(2)) and the safeguards for the new prohibition in Regulation (EU) 2026/1744; 2 August 2026 already applies.","reason":"The calendar did not know 2 December 2026 although the graph had that date, so the countdown widget on other sites showed too late a date.","affected_entity_ids":["praxikon:eu:ai-act:change:2026-12-02-article-50-marking-grace-ends","praxikon:eu:ai-act:change:2026-12-02-new-prohibitions-technical-safeguards"],"affected_paths":["/ai-act-kalender.ics","/embed/ai-act-countdown","/en/ai-act-kalender","/nl/ai-act-kalender"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-en-18286-status-in-copies","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"EN 18286 does not yet provide a presumption of conformity; do not claim harmonised status while the EU citation is absent.","corrected_statement":"Since 14 October 2026 the reference of EN 18286:2026 has been in the Official Journal (Implementing Decision (EU) 2026/2274); applying it gives a presumption of conformity for the requirements it covers.","reason":"The graph processed the decision on 24 October 2026, but four copies kept naming the old state: a monitor update, two source descriptions, a post and the ISO/IEC 42001 object. A reader reported it on 27 October. The monitor update now points to the later update; scripts/check-status-flips.mjs finds such copies on every build since.","affected_entity_ids":["praxikon:eu:ai-act:standard:standard-en-iso-iec-42001-ai-management-system"],"affected_paths":["/en/ai-act-monitor/en-18286-ai-quality-management-standard-published","/en/posts/the-next-ai-act-date-is-2-december-for-most-of-you-it-is-a-supplier-qu","/en/richtsnoeren/nen-en-18286-2026-artificiele-intelligentie-kwaliteitsmanagementsystee","/en/richtsnoeren/nen-en-iso-iec-42001-2026-informatietechnologie-artificiele-intelligen","/en/richtsnoeren/standardisation-of-the-ai-act-policy-page","/nl/ai-act-monitor/en-18286-ai-quality-management-standard-published","/nl/posts/de-volgende-ai-act-datum-is-2-december-voor-de-meesten-is-dat-een-leve","/nl/richtsnoeren/nen-en-18286-2026-artificiele-intelligentie-kwaliteitsmanagementsystee","/nl/richtsnoeren/nen-en-iso-iec-42001-2026-informatietechnologie-artificiele-intelligen","/nl/richtsnoeren/standardisation-of-the-ai-act-policy-page"],"source_url":"https://eur-lex.europa.eu/eli/dec_impl/2026/2274/oj","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-enforcement-tracker-member-states","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"Spain: bill in parliament; Germany: BGBl. 2026 I No. 233; Italy: decrees provisionally approved; Poland: chair expected in October; France: designation removed from the bill.","corrected_statement":"Spain: the bill lapsed on 6 October 2026 when the Cortes were dissolved; Germany: BGBl. 2026 I No. 223; Italy: D.lgs. 160/2026 and 179/2026 published; Poland: chair appointed on 18 September 2026; France: the Senate added the designation on 18 February 2026, the bill is with the Assemblée nationale.","reason":"The enforcement tracker had not been checked against national sources again since 10 August 2026. Read per country on 27 October 2026.","affected_entity_ids":[],"affected_paths":["/en/enforcement","/en/enforcement/france","/en/enforcement/germany","/en/enforcement/italy","/en/enforcement/poland","/en/enforcement/spain","/nl/enforcement","/nl/enforcement/duitsland","/nl/enforcement/frankrijk","/nl/enforcement/italie","/nl/enforcement/polen","/nl/enforcement/spanje"],"source_url":"https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-20742","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-evergreen-dates-and-tense","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"Among others: high-risk systems must be compliant by mid-2026; the FRIA is mandatory from August 2025; enforcement starts on 3 August 2026; first high-risk duties in August 2025.","corrected_statement":"High-risk: 2 December 2027 (Annex III) and 2 August 2028 (Annex I); the FRIA follows from 2 December 2027 and only for certain deployers; Article 50 and GPAI enforcement apply since 2 August 2026; Article 4 requires measures, not a guaranteed level per person.","reason":"Evergreen pages (sector pages, FAQ, complete guide, academy, visual guide) were written before the Digital Omnibus and before 2 August 2026, and did not follow when those dates passed. Found in the check on time-bound wording of 27 October 2026.","affected_entity_ids":[],"affected_paths":["/en/complete-gids-eu-ai-act","/en/faq","/en/sectoren/energie-telecom","/en/sectoren/financiele-diensten","/en/sectoren/gezondheidszorg","/en/sectoren/hr-werkgelegenheid","/en/sectoren/juridische-diensten","/en/sectoren/manufacturing-industrie","/en/sectoren/onderwijs","/en/sectoren/retail-e-commerce","/en/sectoren/technologie-software","/nl/complete-gids-eu-ai-act","/nl/faq","/nl/sectoren/energie-telecom","/nl/sectoren/financiele-diensten","/nl/sectoren/gezondheidszorg","/nl/sectoren/hr-werkgelegenheid","/nl/sectoren/juridische-diensten","/nl/sectoren/manufacturing-industrie","/nl/sectoren/onderwijs","/nl/sectoren/retail-e-commerce","/nl/sectoren/technologie-software"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-gdpr-omnibus-council-mandate-and-edpb-status","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"The Council has no position or mandate yet on the GDPR part of the Digital Omnibus; four EDPB documents are listed as adopted.","corrected_statement":"Since 21 October 2026 the Council has a negotiating mandate (Coreper, ST 14756/26 of 22 October 2026); the Parliament has no position yet and it is not law. Guidelines 1/2024, 01/2025 and 1/2026 and the breach notification template are still consultation versions.","reason":"The mandate followed after the last check of 13 October; the adopted status was an error in the data, since the EDPB publishes only version 1 of these documents.","affected_entity_ids":[],"affected_paths":["/en/posts/the-council-is-negotiating-an-ai-rule-in-the-gdpr-your-balancing-test","/nl/posts/de-raad-onderhandelt-over-een-ai-regel-in-de-avg-uw-belangenafweging-b"],"source_url":"https://data.consilium.europa.eu/doc/document/ST-14756-2026-INIT/en/pdf","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-gpai-code-adequacy","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"Whether the Commission has published its assessment of the adequacy of the GPAI code of practice is not established here.","corrected_statement":"The Commission and the AI Board published on 1 August 2025 their assessment that the code is an adequate voluntary tool; that is not an approval by implementing act.","reason":"The assessment was already in our own source library. Its source label also wrongly attributed general validity within the Union to it.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-56-gpai-codes-of-practice"],"affected_paths":["/en/richtsnoeren/advies-van-de-commissie-over-de-beoordeling-van-de-gpai-gedragscode","/en/verkenner/obligation/article-56-gpai-codes-of-practice","/nl/richtsnoeren/advies-van-de-commissie-over-de-beoordeling-van-de-gpai-gedragscode","/nl/verkenner/obligation/article-56-gpai-codes-of-practice"],"source_url":"https://digital-strategy.ec.europa.eu/en/library/commission-opinion-assessment-general-purpose-ai-code-practice","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-high-risk-guidelines-consultation-closed","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"The draft guidelines of 19 May 2026 are still under consultation; feedback until 23 June 2026.","corrected_statement":"The targeted consultation closed on 23 July 2026, after a four-week extension; the guidelines are not yet final and the Commission plans to adopt them by the end of 2026.","reason":"The extension and the closure were on the Commission's consultation page, but the graph, the source status and the guidelines page did not follow them.","affected_entity_ids":["praxikon:eu:ai-act:guidance:guidance-high-risk-standalone-software-updates-and-add-ons"],"affected_paths":["/en/annex-iii/commission-guidelines-2026","/en/richtsnoeren/ontwerprichtsnoeren-van-de-commissie-over-de-classificatie-van-hoog-ri","/nl/annex-iii/commission-guidelines-2026","/nl/richtsnoeren/ontwerprichtsnoeren-van-de-commissie-over-de-classificatie-van-hoog-ri"],"source_url":"https://digital-strategy.ec.europa.eu/en/consultations/targeted-consultation-draft-guidelines-classification-high-risk-artificial-intelligence-systems","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-nl-supervision-not-designated","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"In the Netherlands the Dutch DPA has been designated as market surveillance authority for the AI Act (elsewhere: RDI designated, or AP and ACM as lead supervisors).","corrected_statement":"No market surveillance authority has been designated yet. Under the cabinet proposal of 20 April 2026 existing supervisors supervise within their domain, the AP becomes the supervisor where none exists, and the AP and RDI coordinate. The implementation act is planned for the second quarter of 2027.","reason":"A proposal was described as a designation in pages, posts and the academy. Found in the check on Dutch status claims of 27 October 2026; verified with the Dutch government and in the Planning Letter on Digital Affairs (Parliamentary Paper 26643, no. 1554).","affected_entity_ids":[],"affected_paths":["/en/boete-calculator","/en/faq","/en/posts/ai-act-supervision-netherlands-who-enforces-what","/en/posts/ireland-first-eu-country-national-ai-act-law","/en/regelgeving/gdpr-en-ai","/nl/boete-calculator","/nl/faq","/nl/posts/ai-privacy","/nl/posts/eu-ai-act-publieke-sector-2025","/nl/posts/euaiact-overview","/nl/posts/ierland-eerste-eu-land-nationale-ai-act-wet","/nl/posts/nederlandse-ai-sandbox-2025","/nl/posts/toezicht-ai-act-nederland-wie-controleert-wat","/nl/regelgeving/gdpr-en-ai"],"source_url":"https://www.rijksoverheid.nl/actueel/nieuws/2026/04/20/kabinet-zet-stap-met-toezicht-op-europese-ai-regels","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-pren-18229-restructure","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"prEN 18229-2 covers accuracy and robustness (Article 15) and prEN 18229-3 transparency and human oversight (Articles 13 and 14); state of June 2026.","corrected_statement":"prEN 18229-2 covers transparency (Article 13), prEN 18229-3 human oversight (Article 14), prEN 18229-4 accuracy and prEN 18229-5 robustness (Article 15). Part 1 (logging) and part 3 are under approval, parts 2, 4 and 5 under drafting.","reason":"JTC 21 restructured the series: part 2 exists as the transparency part since 15 May 2025, part 3 as the oversight part since 11 March 2026, parts 4 and 5 since 24 June 2026. On 8 August 2026 the graph still followed the older division. Found in the standards check of 27 October 2026, read in the CEN-CENELEC work programme.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-40-42-standards-and-specifications","praxikon:eu:ai-act:standard:standard-iso-iec-12792-transparency-taxonomy","praxikon:eu:ai-act:standard:standard-iso-iec-24029-robustness-neural-networks","praxikon:eu:ai-act:standard:standard-pren-18229-1-logging","praxikon:eu:ai-act:standard:standard-pren-18229-2-accuracy-robustness","praxikon:eu:ai-act:standard:standard-pren-18229-3-transparency-human-oversight"],"affected_paths":["/en/verkenner/standard/standard-pren-18229-2-accuracy-robustness","/en/verkenner/standard/standard-pren-18229-3-transparency-human-oversight","/nl/verkenner/standard/standard-pren-18229-2-accuracy-robustness","/nl/verkenner/standard/standard-pren-18229-3-transparency-human-oversight"],"source_url":"https://standards.cencenelec.eu/ords/f?p=205:22:::::FSP_ORG_ID,FSP_LANG_ID:2916257,25","dataset_version":"2.5.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-standardisation-request-m613","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"Standardisation request M/613 was amended by Implementing Decision C(2025)3871 of 23 June 2025.","corrected_statement":"Standardisation request M/613 is Implementing Decision C(2025)3871 of 23 June 2025; it replaced the 2023 request (C(2023)3215).","reason":"Factual error in the prEN 18285 object, found in the standards check of 27 October 2026.","affected_entity_ids":["praxikon:eu:ai-act:standard:standard-pren-18285-conformity-assessment-framework"],"affected_paths":["/en/verkenner/standard/standard-pren-18285-conformity-assessment-framework","/nl/verkenner/standard/standard-pren-18285-conformity-assessment-framework"],"source_url":null,"dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-27-transparency-code-adequacy","corrected_at":"2026-10-27T00:00:00.000Z","severity":"material","previous_statement":"After a positive adequacy assessment, signatories can rely on the code of practice.","corrected_statement":"On 8 July 2026 the Commission concluded that the code adequately covers Article 50(2), (4) and (5); the AI Board adopted its adequacy assessment on 9 July 2026.","reason":"The assessment was published on 9 July 2026, but the graph, an FAQ, the Article 50 guide and two posts still described it as a future condition. The change object gains a later version as of 9 July; the Q&A rendering an addition.","affected_entity_ids":["praxikon:eu:ai-act:change:2026-06-10-transparency-code-of-practice","praxikon:eu:ai-act:guidance:guidance-article-50-code-of-practice-effect"],"affected_paths":["/en/verkenner/guidance/guidance-article-50-code-of-practice-effect","/en/wijzigingen","/nl/verkenner/guidance/guidance-article-50-code-of-practice-effect","/nl/wijzigingen"],"source_url":"https://digital-strategy.ec.europa.eu/en/library/commission-opinion-assessment-code-practice-transparency-ai-generated-content","dataset_version":"2.4.0"},{"id":"praxikon:sys:registry:correction:2026-10-24-en-18286-official-journal-status","corrected_at":"2026-10-24T00:00:00.000Z","severity":"material","previous_statement":"Published by CEN-CENELEC is not the same as cited in the EU Official Journal, and on the official sources consulted that citation cannot yet be found.","corrected_statement":"That citation followed through Implementing Decision (EU) 2026/2274, published on 14 October 2026; since then EN 18286 is a harmonised standard conferring a presumption of conformity for the requirements it covers.","reason":"The source description in the document library dated from before publication in the Official Journal and was not updated for ten days. The graph itself is not an error: there the publication sits as a new version with its own date, beside the version that was right until 14 October 2026.","affected_entity_ids":[],"affected_paths":["/en/richtsnoeren/nen-en-18286-2026-artificiele-intelligentie-kwaliteitsmanagementsystee","/nl/richtsnoeren/nen-en-18286-2026-artificiele-intelligentie-kwaliteitsmanagementsystee"],"source_url":"https://eur-lex.europa.eu/eli/dec_impl/2026/2274/oj","dataset_version":"2.3.0"},{"id":"praxikon:sys:registry:correction:2026-10-20-transparency-code-change-paragraphs","corrected_at":"2026-10-20T00:00:00.000Z","severity":"material","previous_statement":"The Code of Practice on transparency of AI-generated content appeared on 10 June 2026. After a positive adequacy assessment, signatories can rely on it for Article 50(2), (3) and (5). Section 1 addresses providers, section 2 deployers.","corrected_statement":"The Code of Practice on transparency of AI-generated content appeared on 10 June 2026. After a positive adequacy assessment, signatories can rely on it for Article 50(2), (4) and (5). Section 1 addresses providers, section 2 deployers.","reason":"The correction of 5 October 2026 on the same paragraphs repaired the blog posts but missed this statement in the graph. Article 50(3) concerns emotion recognition and biometric categorisation, paragraph 4 deep fakes. Paragraph 7, as replaced by Regulation (EU) 2026/1744, concerns codes of practice for the detection, marking and labelling of artificially generated or manipulated content and names paragraphs 2 and 4.","affected_entity_ids":["praxikon:eu:ai-act:change:2026-06-10-transparency-code-of-practice"],"affected_paths":["/en/verkenner/change/2026-06-10-transparency-code-of-practice","/en/wijzigingen","/nl/verkenner/change/2026-06-10-transparency-code-of-practice","/nl/wijzigingen"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.2.2"},{"id":"praxikon:sys:registry:correction:2026-10-11-article-86-systems-already-in-use","corrected_at":"2026-10-11T00:00:00.000Z","severity":"material","previous_statement":"Decision D1 chooses the text reading, so Article 86 applies from 2 August 2026 to systems already in use at that point.","corrected_statement":"Article 86 formally applies from 2 August 2026, but in practice can only be invoked once the Annex III regime operates, on 2 December 2027. For a system already in use before then, Article 111(2) applies: only on a significant change in design, and for public authorities by 2 August 2030 at the latest.","reason":"Article 111(2) governs when the Regulation applies to high-risk systems already in use; the note left that out. The application date in deadline_at did not move.","affected_entity_ids":["praxikon:eu:ai-act:obligation:article-86-right-to-explanation"],"affected_paths":["/en/verplichtingen/article-86-right-to-explanation","/nl/verplichtingen/article-86-right-to-explanation"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.2.1"},{"id":"praxikon:sys:registry:correction:2026-10-10-dpia-fria-fines","corrected_at":"2026-10-10T14:46:18.000Z","severity":"material","previous_statement":"Non-compliance with the FRIA duty can lead to fines of up to 35 million euros or 7% of worldwide annual turnover (in the DPIA guide: up to 3% or 15 million euros); DPIA violations under the GDPR up to 4% of worldwide turnover.","corrected_statement":"Article 27 AI Act is not in the fines list of Article 99(4) (up to 15 million euros or 3%); penalties for the FRIA follow from the national rules Member States lay down under Article 99(1). 35 million euros or 7% applies only to prohibited practices (Article 99(3)). A missing or deficient DPIA falls under Article 83(4) GDPR: up to 10 million euros or 2%.","reason":"Article 99(3) AI Act covers only Article 5. Article 99(4) names Articles 16, 22, 23, 24, 25(2) and (4), 26, 31, 33, 34 and 50, not Article 27. Article 35 GDPR falls under Article 83(4)(a) GDPR. Repaired on 10 October 2026 when the DPIA pages were merged into one owner per language; recorded in this register on 24 October 2026.","affected_entity_ids":[],"affected_paths":["/en/dpia-vs-fria","/en/posts/dpia-ai-systems-when-required-guide","/nl/dpia-vs-fria","/nl/posts/dpia-ai-systemen-wanneer-verplicht-handleiding","/nl/posts/dpia-vs-fria-praktische-vergelijking"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-10-fria-scope-duration-consultation","corrected_at":"2026-10-10T14:46:18.000Z","severity":"material","previous_statement":"Not all high-risk AI systems require a FRIA; exempted are safety components of products, critical infrastructure and AI for fraud detection. A FRIA takes 4 to 8 weeks on average and affected parties such as works councils must be consulted.","corrected_statement":"A FRIA applies only to high-risk systems listed in Annex III (Article 6(2)), not to point 2 (critical infrastructure), and only to bodies governed by public law, private entities providing public services and deployers under Annex III point 5(b) and (c). AI used only to detect financial fraud falls outside point 5(b); that is not a general exemption from the FRIA. Article 27 sets no duration and no duty to consult affected parties.","reason":"Article 27(1) and Annex III point 5(b) AI Act. The duration and the duty to consult had no source. Repaired on 10 October 2026 when the DPIA pages were merged into one owner per language; recorded in this register on 24 October 2026.","affected_entity_ids":[],"affected_paths":["/en/dpia-vs-fria","/en/posts/dpia-vs-fria-practical-comparison","/nl/dpia-vs-fria","/nl/posts/dpia-vs-fria-praktische-vergelijking"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-05-code-of-practice-paragraphs","corrected_at":"2026-10-05T18:00:00.000Z","severity":"material","previous_statement":"Signatories of the code of practice can rely on it for Article 50(2), (3) and (5).","corrected_statement":"After a positive adequacy assessment by the Commission, signatories can rely on it for Article 50(2), (4) and (5).","reason":"Article 50(7) names paragraphs 2 and 4; the guidelines (points 146 to 148) and the Commission page name (2), (4) and (5). Paragraph 3 (emotion recognition) is outside the code.","affected_entity_ids":[],"affected_paths":["/en/posts/article-50-applies-now-what-changed","/en/posts/article-50-became-enforceable-on-sunday-here-is-the-new-baseline","/en/posts/one-more-ai-act-deadline-falls-before-2-august-27-july-1800","/nl/posts/artikel-50-geldt-nu-wat-er-veranderde","/nl/posts/er-valt-nog-een-ai-act-deadline-voor-2-augustus-27-juli-1800","/nl/posts/sinds-zondag-is-artikel-50-handhaafbaar-dit-is-de-nieuwe-stand"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-05-code-of-practice-status","corrected_at":"2026-10-05T18:00:00.000Z","severity":"material","previous_statement":"The code of practice on AI content transparency is still a draft or forthcoming.","corrected_statement":"The final, voluntary code of practice was published on 10 June 2026; the final Article 50 guidelines on 20 July 2026.","reason":"The posts were written before publication and had not been updated.","affected_entity_ids":[],"affected_paths":["/en/posts/article-50-practical-labeling-detection","/en/posts/article-50-provider-deployer-transparency-eu-ai-act","/en/posts/article-50-transparency-checklist-2-august-2026","/en/posts/article-50-transparency-deadline-2-august-2026","/en/posts/code-of-practice-transparency-ai-content","/en/posts/deepfakes-transparency-ai-act-2026","/nl/posts/artikel-50-praktijk-labeling-detectie","/nl/posts/artikel-50-provider-deployer-transparantie-eu-ai-act","/nl/posts/artikel-50-transparantie-checklist-2-augustus-2026","/nl/posts/artikel-50-transparantie-deadline-2-augustus-2026","/nl/posts/code-of-practice-transparantie-ai-content","/nl/posts/deepfakes-transparantie-ai-act-2026"],"source_url":"https://digital-strategy.ec.europa.eu/en/policies/code-practice-ai-generated-content","dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-05-deployer-labelling-scope","corrected_at":"2026-10-05T18:00:00.000Z","severity":"material","previous_statement":"Deployers must visibly label AI-generated or manipulated content; anyone using AI images or video in campaigns or course material has marking and disclosure duties.","corrected_statement":"A deployer only has a disclosure duty for deep fakes and for AI text published to inform the public on matters of public interest without human review and editorial responsibility. Machine-readable marking is a provider duty.","reason":"Article 50(2) and (4) and the guidelines of 20 July 2026, points 113 to 116 and 131 to 138.","affected_entity_ids":[],"affected_paths":["/en/posts/ai-act-training-providers-four-steps","/en/posts/article-50-practical-labeling-detection","/en/posts/article-50-provider-deployer-transparency-eu-ai-act","/en/posts/article-50-transparency-checklist-2-august-2026","/en/posts/code-of-practice-transparency-ai-content","/nl/posts/ai-act-opleidingsorganisaties-vier-stappen","/nl/posts/artikel-50-praktijk-labeling-detectie","/nl/posts/artikel-50-provider-deployer-transparantie-eu-ai-act","/nl/posts/artikel-50-transparantie-checklist-2-augustus-2026","/nl/posts/code-of-practice-transparantie-ai-content"],"source_url":"https://digital-strategy.ec.europa.eu/en/library/guidelines-transparency-obligations-providers-and-deployers-ai-systems","dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-05-emotion-recognition-scope","corrected_at":"2026-10-05T18:00:00.000Z","severity":"material","previous_statement":"Emotion recognition at work and in education is absolutely prohibited, including text analysis; transparency follows from Articles 86 and 87.","corrected_statement":"The ban has a medical or safety exception and requires biometric data; text analysis falls outside. Outside the ban, emotion recognition is high-risk (Annex III point 1(c), from 2 December 2027) and the Article 50(3) information duty applies.","reason":"Article 3(39), Article 5(1)(f) and Article 50(3).","affected_entity_ids":[],"affected_paths":["/en/posts/ai-act-customer-contact-ai-chatbots-compliance","/en/posts/ap-emotion-recognition-report-2025","/nl/posts/ai-act-klantcontact-ai-chatbots-compliance","/nl/posts/ap-emotieherkenning-rapport-2025"],"source_url":"https://eur-lex.europa.eu/eli/reg/2024/1689/oj","dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-05-fines-smaller-undertakings","corrected_at":"2026-10-05T18:00:00.000Z","severity":"material","previous_statement":"Fine of up to EUR 15 million or 3 percent of worldwide annual turnover, whichever is higher, without exception.","corrected_statement":"The higher amount applies, except for SMEs, start-ups and small mid-caps, for whom the lower applies.","reason":"Article 99(6) and (6a), inserted by Regulation (EU) 2026/1744.","affected_entity_ids":[],"affected_paths":["/en/posts/article-50-applies-now-what-changed","/en/posts/article-50-became-enforceable-on-sunday-here-is-the-new-baseline","/en/posts/article-50-enforcement-fines-ai-act-2026","/en/posts/article-50-transparency-obligations-practical-2026","/en/posts/chatbot-ai-disclosure-ai-act-2026","/en/posts/deepfakes-transparency-ai-act-2026","/en/posts/emotion-recognition-biometrics-transparency-ai-act-2026","/nl/posts/artikel-50-geldt-nu-wat-er-veranderde","/nl/posts/artikel-50-handhaving-boetes-ai-act-2026","/nl/posts/artikel-50-transparantie-verplichtingen-praktisch-2026","/nl/posts/chatbot-ai-kenbaar-maken-ai-act-2026","/nl/posts/deepfakes-transparantie-ai-act-2026","/nl/posts/emotieherkenning-biometrie-transparantie-ai-act-2026","/nl/posts/sinds-zondag-is-artikel-50-handhaafbaar-dit-is-de-nieuwe-stand"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-05-nl-supervisors-designation","corrected_at":"2026-10-05T18:00:00.000Z","severity":"material","previous_statement":"In the Netherlands the supervisory authorities have been designated, with the Dutch DPA as coordinator.","corrected_statement":"The Netherlands has not yet formally designated any supervisor by law; the bill intends the Dutch DPA and the RDI as coordinators.","reason":"The bill for the Dutch AI Act implementation law has not yet been submitted.","affected_entity_ids":[],"affected_paths":["/en/posts/ai-act-customer-contact-ai-chatbots-compliance","/en/posts/article-50-became-enforceable-on-sunday-here-is-the-new-baseline","/nl/posts/ai-act-klantcontact-ai-chatbots-compliance","/nl/posts/sinds-zondag-is-artikel-50-handhaafbaar-dit-is-de-nieuwe-stand"],"source_url":null,"dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-05-omnibus-status-and-transition","corrected_at":"2026-10-05T18:00:00.000Z","severity":"material","previous_statement":"The Digital Omnibus agreement still has to be formally adopted; the 2 December 2026 date was not granted by the Omnibus; Article 4 is unchanged.","corrected_statement":"Regulation (EU) 2026/1744 has been in force since 27 July 2026. It added the transition to 2 December 2026 for the paragraph 2 marking (Article 111(4)) and reworded Article 4 as a duty to take measures.","reason":"Status claims dated from before publication in the Official Journal.","affected_entity_ids":[],"affected_paths":["/en/posts/article-50-transparency-deadline-2-august-2026","/nl/posts/artikel-50-transparantie-deadline-2-augustus-2026"],"source_url":"https://eur-lex.europa.eu/eli/reg/2026/1744/oj","dataset_version":"2.2.0"},{"id":"praxikon:sys:registry:correction:2026-10-05-provider-test-and-obvious-exception","corrected_at":"2026-10-05T18:00:00.000Z","severity":"material","previous_statement":"Whether you become the provider follows from Article 25; a label such as digital assistant may meet the obviousness exception.","corrected_statement":"Outside high-risk, provider status follows from Article 3(3) (guidelines point 11). The exception is narrow: almost no doubt, and a generic label such as assistant is not enough.","reason":"Article 25 governs high-risk systems; guidelines points 11, 38, 44 and 45.","affected_entity_ids":[],"affected_paths":["/en/posts/article-50-field-test-dutch-chatbots","/en/posts/article-50-transparency-checklist-2-august-2026","/en/posts/article-50-transparency-deadline-2-august-2026","/en/posts/article-50-transparency-obligations-practical-2026","/en/posts/chatbot-ai-disclosure-ai-act-2026","/nl/posts/artikel-50-praktijktest-nederlandse-chatbots","/nl/posts/artikel-50-transparantie-checklist-2-augustus-2026","/nl/posts/artikel-50-transparantie-deadline-2-augustus-2026","/nl/posts/artikel-50-transparantie-verplichtingen-praktisch-2026","/nl/posts/chatbot-ai-kenbaar-maken-ai-act-2026"],"source_url":"https://digital-strategy.ec.europa.eu/en/library/guidelines-transparency-obligations-providers-and-deployers-ai-systems","dataset_version":"2.2.0"}],"links":{"self":"https://www.praxikon.com/api/v1/corrections?lang=en","licence":"https://www.praxikon.com/nl/legal/terms","human_page":"https://www.praxikon.com/en/correcties","methodology":"https://www.praxikon.com/en/methodologie"}}