Skip to content

Regulation Tracker / EU AI Act / Chapter XIII

Article 112
Evaluation and review

Chapter XIII — Amendment to Regulation (EC) No 300/2008. 8 distinct duties, powers or definitions are coded from this article, applying from 2026-08-02. 1 reaches legal practice directly.

official text: EUR-Lex (consolidated 27 Jul 2026) · text © European Union (Decision 2011/833/EU) · coding CC BY 4.0 · data: obligations.json · Hugging Face · GitHub

The text

1. The Commission shall assess the need for amendment of the list set out in Annex III and of the list of prohibited AI practices laid down in Article 5, once a year following the entry into force of this Regulation, and until the end of the period of the delegation of power laid down in Article 97. The Commission shall submit the findings of that assessment to the European Parliament and the Council.

2. By 2 August 2028 and every four years thereafter, the Commission shall evaluate and report to the European Parliament and to the Council on the following:

  1. (a)the need for amendments extending existing area headings or adding new area headings in Annex III;
  2. (b)amendments to the list of AI systems requiring additional transparency measures in Article 50;
  3. (c)amendments enhancing the effectiveness of the supervision and governance system.

3. By 2 August 2029 and every four years thereafter, the Commission shall submit a report on the evaluation and review of this Regulation to the European Parliament and to the Council. The report shall include an assessment with regard to the structure of enforcement and the possible need for a Union agency to resolve any identified shortcomings. On the basis of the findings, that report shall, where appropriate, be accompanied by a proposal for amendment of this Regulation. The reports shall be made public.

4. The reports referred to in paragraph 2 shall pay specific attention to the following:

  1. (a)the status of the financial, technical and human resources of the national competent authorities in order to effectively perform the tasks assigned to them under this Regulation;
  2. (b)the state of penalties, in particular administrative fines as referred to in Article 99(1), applied by Member States for infringements of this Regulation;
  3. (c)adopted harmonised standards and common specifications developed to support this Regulation;
  4. (d)the number of undertakings that enter the market after the entry into application of this Regulation, and how many of them are SMEs.

5. By 2 August 2028, the Commission shall evaluate the functioning of the AI Office, whether the AI Office has been given sufficient powers and competences to fulfil its tasks, and whether it would be relevant and needed for the proper implementation and enforcement of this Regulation to upgrade the AI Office and its enforcement competences and to increase its resources. The Commission shall submit a report on its evaluation to the European Parliament and to the Council.

6. By 2 August 2028 and every four years thereafter, the Commission shall submit a report on the review of the progress on the development of standardisation deliverables on the energy-efficient development of general-purpose AI models, and asses the need for further measures or actions, including binding measures or actions. The report shall be submitted to the European Parliament and to the Council, and it shall be made public.

7. By 2 August 2028 and every three years thereafter, the Commission shall evaluate the impact and effectiveness of voluntary codes of conduct to foster the application of the requirements set out in Chapter III, Section 2 for AI systems other than high-risk AI systems and possibly other additional requirements for AI systems other than high-risk AI systems, including as regards environmental sustainability.

8. For the purposes of paragraphs 1 to 7, the Board, the Member States and national competent authorities shall provide the Commission with information upon its request and without undue delay.

9. In carrying out the evaluations and reviews referred to in paragraphs 1 to 7, the Commission shall take into account the positions and findings of the Board, of the European Parliament, of the Council, and of other relevant bodies or sources.

10. The Commission shall, if necessary, submit appropriate proposals to amend this Regulation, in particular taking into account developments in technology, the effect of AI systems on health and safety, and on fundamental rights, and in light of the state of progress in the information society.

11. To guide the evaluations and reviews referred to in paragraphs 1 to 7 of this Article, the AI Office shall undertake to develop an objective and participative methodology for the evaluation of risk levels based on the criteria outlined in the relevant Articles and the inclusion of new systems in:

  1. (a)the list set out in Annex III, including the extension of existing area headings or the addition of new area headings in that Annex;
  2. (b)the list of prohibited practices set out in Article 5; and
  3. (c)the list of AI systems requiring additional transparency measures pursuant to Article 50.

12. Any amendment to this Regulation pursuant to paragraph 10, or relevant delegated or implementing acts, which concerns sectoral Union harmonisation legislation listed in Section B of Annex I shall take into account the regulatory specificities of each sector, and the existing governance, conformity assessment and enforcement mechanisms and authorities established therein.

13. By 2 August 2031, the Commission shall carry out an assessment of the enforcement of this Regulation and shall report on it to the European Parliament, the Council and the European Economic and Social Committee, taking into account the first years of application of this Regulation. On the basis of the findings, that report shall, where appropriate, be accompanied by a proposal for amendment of this Regulation with regard to the structure of enforcement and the need for a Union agency to resolve any identified shortcomings.

What it requires, coded

SafeLegalAI's reading of each duty in this article: who, what, from when, under which fine tier, and whether it reaches a firm, chambers, court or legal-AI vendor. Descriptive, not advice; the quoted words are the Regulation's.

  1. Art. 112(1)Governance institutionalNot tier-specificlegal practice

    The Commission must annually assess the need to amend Annex III and Article 5 prohibited-practice lists until the Article 97 delegation period ends, and submit findings to Parliament and Council.

    "The Commission shall assess the need for amendment of the list set out in Annex III and of the list of prohibited AI practices laid down in Article 5, once a year"
    Who
    Commission
    From
    (Art. 113 second subparagraph)
    Legal practice
    Reaches legal practice directly — Annex III point 8 and Article 5 changes could materially affect legal-AI providers, courts and deployers.
    See
    Annex III · Art. 5 · Art. 97
  2. Art. 112(2)Governance institutionalNot tier-specific

    By 2 August 2028 and every four years thereafter, the Commission must evaluate and report on Annex III headings, Article 50 transparency systems and supervision/governance effectiveness.

    "By 2 August 2028 and every four years thereafter, the Commission shall evaluate and report to the European Parliament and to the Council"
    Who
    Commission
    From
    (Art. 113 second subparagraph)
    Legal practice
    Reaches legal-AI vendors or public bodies — Reviews may change high-risk and transparency categories relevant to legal-sector AI.
    See
    Annex III · Art. 50
  3. Art. 112(3)Governance institutionalNot tier-specific

    By 2 August 2029 and every four years thereafter, the Commission must report on evaluation and review of the Regulation, including enforcement structure and possible need for a Union agency.

    "By 2 August 2029 and every four years thereafter, the Commission shall submit a report on the evaluation and review of this Regulation"
    Who
    Commission
    From
    (Art. 113 second subparagraph)
    Legal practice
    Institutional — This is a general review duty that may affect future AI Act governance.
  4. Art. 112(4)Governance institutionalNot tier-specific

    Article 112 reports must pay specific attention to authority resources, penalties, harmonised standards/common specifications and numbers of new market entrants and SMEs.

    "The reports referred to in paragraph 2 shall pay specific attention to the following"
    Who
    Commission
    From
    (Art. 113 second subparagraph)
    Legal practice
    Reaches legal-AI vendors or public bodies — Penalty and standards data may affect legal-AI compliance expectations.
    See
    Art. 99 · Art. 40 · Art. 41
  5. Art. 112(5)Governance institutionalNot tier-specific

    By 2 August 2028, the Commission must evaluate the AI Office’s functioning, powers, competences and resources and report to Parliament and Council.

    "By 2 August 2028, the Commission shall evaluate the functioning of the AI Office"
    Who
    Commission
    From
    (Art. 113 second subparagraph)
    Legal practice
    Institutional — This affects institutional enforcement capacity rather than direct legal-sector duties.
    See
    AI Office
  6. Art. 112(6)Governance institutionalGeneral-purpose AI

    By 2 August 2028 and every four years thereafter, the Commission must report publicly on standardisation deliverables for energy-efficient GPAI model development and assess whether further measures are needed.

    "the Commission shall submit a report on the review of the progress on the development of standardisation deliverables on the energy-efficient development of general-purpose AI models"
    Who
    Commission
    From
    (Art. 113 second subparagraph)
    Legal practice
    Institutional — This is most relevant to GPAI model providers, including models used in legal AI.
    See
    general-purpose AI models
  7. Art. 112(7)Governance institutionalAll AI systems

    By 2 August 2028 and every three years thereafter, the Commission must evaluate the impact and effectiveness of voluntary codes of conduct for non-high-risk AI requirements and sustainability.

    "the Commission shall evaluate the impact and effectiveness of voluntary codes of conduct"
    Who
    Commission
    From
    (Art. 113 second subparagraph)
    Legal practice
    Reaches legal-AI vendors or public bodies — Voluntary codes may influence legal-sector AI governance for non-high-risk systems.
    See
    Art. 95 · Chapter III Section 2
  8. Art. 112(8)Cooperation with authoritiesNot tier-specific

    The Board, Member States and national authorities must provide requested information without undue delay, and the AI Office must develop a participative methodology for evaluating risk levels and list additions.

    "the Board, the Member States and national competent authorities shall provide the Commission with information upon its request and without undue delay"
    Who
    AI Board, Member State, National competent authority, AI Office, Commission
    From
    (Art. 113 second subparagraph)
    Legal practice
    Reaches legal-AI vendors or public bodies — Risk-level methodology can affect whether legal AI becomes listed as high-risk or transparency-risk.
    See
    Art. 112(11) · Annex III · Art. 5 · Art. 50

Cite

Regulation (EU) 2024/1689, Article 112 (Evaluation and review), as amended by Regulation (EU) 2026/1744, consolidated text of 27 July 2026, https://eur-lex.europa.eu/eli/reg/2024/1689/2026-07-27/eng#art_112 — text © European Union; only the Official Journal is authentic. Coding: SafeLegalAI (published by Cognesio LLP), "EU AI Act, structured", safelegalai.com/regulation/eu-ai-act/article-112, accessed 2026-09-08, CC BY 4.0.