# Human Oversight — legal context bundle > Curated from overview.legal on 2026-08-22. Canonical page: https://overview.legal/topics/human-oversight-ai > Sources are cited per item. Verify against the official texts before relying on them. This new topic is needed because human oversight is a specific and distinct requirement under the AI Act that deserves dedicated coverage, encompassing mechanisms for human control, intervention, and review of AI system operations and decisions. ## Overview ## Legal Framework Article 14 of the AI Act establishes human oversight as a core obligation for high-risk AI systems. The provision requires that such systems be designed and developed so that they can be effectively overseen by natural persons during the period in which they are in use. Oversight must be proportionate to the risks, level of autonomy, and context of use of the system. The rationale, reinforced by Recital 72, is to counteract the opacity and complexity that characterise many AI systems. High-risk systems must enable deployers to understand how outputs are generated, evaluate system functionality, and comprehend inherent strengths and limitations. Human oversight is not a passive monitoring duty — it requires that individuals have the capacity to understand, interpret, and, where necessary, intervene in or override system outputs. Article 14 identifies three functions that oversight must serve: preventing or minimising risks to health, safety, or fundamental rights during use; ensuring that outputs do not undermine those protected interests; and enabling the cessation of system operation through a "stop" mechanism where risks materialise. Where feasible, oversight measures must be built directly into the system by the provider, rather than left entirely to deployers. ## Key Developments Because the AI Act entered into force only in August 2024, no enforcement decisions or case law interpreting Article 14 have yet emerged. However, the provision builds on well-established principles from GDPR enforcement that treat meaningful human review as a substantive, not merely formal, requirement. Dutch DPA guidance on automated decision-making under Article 22 GDPR has consistently held that human involvement must be genuine — reviewers must have the authority and competence to override automated outputs. The same logic will animate Article 14 enforcement: tokenistic oversight, where a human rubber-stamps AI recommendations without independent assessment, will not satisfy the requirement. The AI Office and national competent authorities are expected to issue further guidance on what constitutes "effective" oversight before the high-risk obligations apply in August 2026. ## Practical Guidance - **Design for intervention from the outset**: Providers must build oversight mechanisms into high-risk AI systems before market placement, including interfaces that allow deployers to monitor operation and interpret outputs in real time. - **Equip oversight personnel with genuine authority**: Designate individuals who have both the technical competence to understand system outputs and the organisational authority to override or halt them — symbolic approval will not suffice. - **Implement a functional stop mechanism**: Ensure that the system can be immediately taken out of service when risks to health, safety, or fundamental rights materialise, as Article 14(4)(c) requires. - **Document the oversight architecture**: Maintain records showing how human oversight is structured, who is responsible, what information they receive, and how intervention decisions are made and logged. - **Align with instructions for use**: Providers' accompanying documentation must clearly explain the oversight capabilities built into the system, enabling deployers to fulfil their own oversight obligations effectively. ## Legislation (full text of key provisions) ### Human oversight *Source: AI Act, aiact-art-14-en, 2024-06-12 — https://overview.legal/posts/92201* ### Recital 73 — human oversight of high-risk AI *Source: AI Act, aiact-rec-73-en, 2024-06-12 — https://overview.legal/posts/93828* High-risk AI systems should be designed and developed in such a way that natural persons can oversee their functioning, ensure that they are used as intended and that their impacts are addressed over the system’s lifecycle. To that end, appropriate human oversight measures should be identified by the provider of the system before its placing on the market or putting into service. In particular, where appropriate, such measures should guarantee that the system is subject to in-built operational constraints that cannot be overridden by the system itself and is responsive to the human operator, and that the natural persons to whom human oversight has been assigned have the necessary competence, training and authority to carry out that role. It is also essential, as appropriate, to ensure that high-risk AI systems include mechanisms to guide and inform a natural person to whom human oversight has been assigned to make informed decisions if, when and how to intervene in order to avoid negative consequences or risks, or stop the system if it does not perform as intended. Considering the significant consequences for persons in the case of an incorrect match by certain biometric identification systems, it is appropriate to provide for an enhanced human oversight requirement for those systems so that no action or decision may be taken by the deployer on the basis of the identification resulting from the system unless this has been separately verified and confirmed by at least two natural persons. Those persons could be from one or more entities and include the person operating or using the system. This requirement should not pose unnecessary burden or delays and it could be sufficient that the separate verifications by the different persons are automatically recorded in the logs generated by the system. Given the specificities of the areas of law enforcement, migration, border control and asylum, this requirement should not apply where Union or national law considers the application of that requirement to be disproportionate. ### Recital 72 — transparency requirements for high-risk AI systems *Source: AI Act, aiact-rec-72-en, 2024-06-12 — https://overview.legal/posts/93826* To address concerns related to opacity and complexity of certain AI systems and help deployers to fulfil their obligations under this Regulation, transparency should be required for high-risk AI systems before they are placed on the market or put it into service. High-risk AI systems should be designed in a manner to enable deployers to understand how the AI system works, evaluate its functionality, and comprehend its strengths and limitations. High-risk AI systems should be accompanied by appropriate information in the form of instructions of use. Such information should include the characteristics, capabilities and limitations of performance of the AI system. Those would cover information on possible known and foreseeable circumstances related to the use of the high-risk AI system, including deployer action that may influence system behaviour and performance, under which the AI system can lead to risks to health, safety, and fundamental rights, on the changes that have been pre-determined and assessed for conformity by the provider and on the relevant human oversight measures, including the measures to facilitate the interpretation of the outputs of the AI system by the deployers. Transparency, including the accompanying instructions for use, should assist deployers in the use of the system and support informed decision making by them. Deployers should, inter alia, be in a better position to make the correct choice of the system that they intend to use in light of the obligations applicable to them, be educated about the intended and precluded uses, and use the AI system correctly and as appropriate. In order to enhance legibility and accessibility of the information included in the instructions of use, where appropriate, illustrative examples, for instance on the limitations and on the intended and precluded uses of the AI system, should be included. Providers should ensure that all documentation, including the instructions for use, contains meaningful, comprehensive, accessible and understandable information, taking into account the needs and foreseeable knowledge of the target deployers. Instructions for use should be made available in a language which can be easily understood by target deployers, as determined by the Member State concerned. ### Recital 66 — risk management requirements for high-risk AI *Source: AI Act, aiact-rec-66-en, 2024-06-12 — https://overview.legal/posts/93814* Requirements should apply to high-risk AI systems as regards risk management, the quality and relevance of data sets used, technical documentation and record-keeping, transparency and the provision of information to deployers, human oversight, and robustness, accuracy and cybersecurity. Those requirements are necessary to effectively mitigate the risks for health, safety and fundamental rights. As no other less trade restrictive measures are reasonably available those requirements are not unjustified restrictions to trade. ### Recital 91 — deployer responsibilities for high-risk AI systems *Source: AI Act, aiact-rec-91-en, 2024-06-12 — https://overview.legal/posts/93864* Given the nature of AI systems and the risks to safety and fundamental rights possibly associated with their use, including as regards the need to ensure proper monitoring of the performance of an AI system in a real-life setting, it is appropriate to set specific responsibilities for deployers. Deployers should in particular take appropriate technical and organisational measures to ensure they use high-risk AI systems in accordance with the instructions of use and certain other obligations should be provided for with regard to monitoring of the functioning of the AI systems and with regard to record-keeping, as appropriate. Furthermore, deployers should ensure that the persons assigned to implement the instructions for use and human oversight as set out in this Regulation have the necessary competence, in particular an adequate level of AI literacy, training and authority to properly fulfil those tasks. Those obligations should be without prejudice to other deployer obligations in relation to high-risk AI systems under Union or national law. ### Recital 12 — AI system definition and characteristics *Source: AI Act, aiact-rec-12-en, 2024-06-12 — https://overview.legal/posts/93706* The notion of ‘AI system’ in this Regulation should be clearly defined and should be closely aligned with the work of international organisations working on AI to ensure legal certainty, facilitate international convergence and wide acceptance, while providing the flexibility to accommodate the rapid technological developments in this field. Moreover, the definition should be based on key characteristics of AI systems that distinguish it from simpler traditional software systems or programming approaches and should not cover systems that are based on the rules defined solely by natural persons to automatically execute operations. A key characteristic of AI systems is their capability to infer. This capability to infer refers to the process of obtaining the outputs, such as predictions, content, recommendations, or decisions, which can influence physical and virtual environments, and to a capability of AI systems to derive models or algorithms, or both, from inputs or data. The techniques that enable inference while building an AI system include machine learning approaches that learn from data how to achieve certain objectives, and logic- and knowledge-based approaches that infer from encoded knowledge or symbolic representation of the task to be solved. The capacity of an AI system to infer transcends basic data processing by enabling learning, reasoning or modelling. The term ‘machine-based’ refers to the fact that AI systems run on machines. The reference to explicit or implicit objectives underscores that AI systems can operate according to explicit defined objectives or to implicit objectives. The objectives of the AI system may be different from the intended purpose of the AI system in a specific context. For the purposes of this Regulation, environments should be understood to be the contexts in which the AI systems operate, whereas outputs generated by the AI system reflect different functions performed by AI systems and include predictions, content, recommendations or decisions. AI systems are designed to operate with varying levels of autonomy, meaning that they have some degree of independence of actions from human involvement and of capabilities to operate without human intervention. The adaptiveness that an AI system could exhibit after deployment, refers to self-learning capabilities, allowing the system to change while in use. AI systems can be used on a stand-alone basis or as a component of a product, irrespective of whether the system is physically integrated into the product (embedded) or serves the functionality of the product without being integrated therein (non-embedded). ### Recital 96 — fundamental rights impact assessment deployers *Source: AI Act, aiact-rec-96-en, 2024-06-12 — https://overview.legal/posts/93874* In order to efficiently ensure that fundamental rights are protected, deployers of high-risk AI systems that are bodies governed by public law, or private entities providing public services and deployers of certain high-risk AI systems listed in an annex to this Regulation, such as banking or insurance entities, should carry out a fundamental rights impact assessment prior to putting it into use. Services important for individuals that are of public nature may also be provided by private entities. Private entities providing such public services are linked to tasks in the public interest such as in the areas of education, healthcare, social services, housing, administration of justice. The aim of the fundamental rights impact assessment is for the deployer to identify the specific risks to the rights of individuals or groups of individuals likely to be affected, identify measures to be taken in the case of a materialisation of those risks. The impact assessment should be performed prior to deploying the high-risk AI system, and should be updated when the deployer considers that any of the relevant factors have changed. The impact assessment should identify the deployer’s relevant processes in which the high-risk AI system will be used in line with its intended purpose, and should include a description of the period of time and frequency in which the system is intended to be used as well as of specific categories of natural persons and groups who are likely to be affected in the specific context of use. The assessment should also include the identification of specific risks of harm likely to have an impact on the fundamental rights of those persons or groups. While performing this assessment, the deployer should take into account information relevant to a proper assessment of the impact, including but not limited to the information given by the provider of the high-risk AI system in the instructions for use. In light of the risks identified, deployers should determine measures to be taken in the case of a materialisation of those risks, including for example governance arrangements in that specific context of use, such as arrangements for human oversight according to the instructions of use or, complaint handling and redress procedures, as they could be instrumental in mitigating risks to fundamental rights in concrete use-cases. After performing that impact assessment, the deployer should notify the relevant market surveillance authority. Where appropriate, to collect relevant information necessary to perform the impact assessment, deployers of high-risk AI system, in particular when AI systems are used in the public sector, could involve relevant stakeholders, including the representatives of groups of persons likely to be affected by the AI system, independent experts, and civil society organisations in conducting such impact assessments and designing measures to be taken in the case of materialisation of the risks. The European Artificial Intelligence Office (AI Office) should develop a template for a questionnaire in order to facilitate compliance and reduce the administrative burden for deployers. ### Recital 53 — low risk AI systems clarification *Source: AI Act, aiact-rec-53-en, 2024-06-12 — https://overview.legal/posts/93788* It is also important to clarify that there may be specific cases in which AI systems referred to in pre-defined areas specified in this Regulation do not lead to a significant risk of harm to the legal interests protected under those areas because they do not materially influence the decision-making or do not harm those interests substantially. For the purposes of this Regulation, an AI system that does not materially influence the outcome of decision-making should be understood to be an AI system that does not have an impact on the substance, and thereby the outcome, of decision-making, whether human or automated. An AI system that does not materially influence the outcome of decision-making could include situations in which one or more of the following conditions are fulfilled. The first such condition should be that the AI system is intended to perform a narrow procedural task, such as an AI system that transforms unstructured data into structured data, an AI system that classifies incoming documents into categories or an AI system that is used to detect duplicates among a large number of applications. Those tasks are of such narrow and limited nature that they pose only limited risks which are not increased through the use of an AI system in a context that is listed as a high-risk use in an annex to this Regulation. The second condition should be that the task performed by the AI system is intended to improve the result of a previously completed human activity that may be relevant for the purposes of the high-risk uses listed in an annex to this Regulation. Considering those characteristics, the AI system provides only an additional layer to a human activity with consequently lowered risk. That condition would, for example, apply to AI systems that are intended to improve the language used in previously drafted documents, for example in relation to professional tone, academic style of language or by aligning text to a certain brand messaging. The third condition should be that the AI system is intended to detect decision-making patterns or deviations from prior decision-making patterns. The risk would be lowered because the use of the AI system follows a previously completed human assessment which it is not meant to replace or influence, without proper human review. Such AI systems include for instance those that, given a certain grading pattern of a teacher, can be used to check ex post whether the teacher may have deviated from the grading pattern so as to flag potential inconsistencies or anomalies. The fourth condition should be that the AI system is intended to perform a task that is only preparatory to an assessment relevant for the purposes of the AI systems listed in an annex to this Regulation, thus making the possible impact of the output of the system very low in terms of representing a risk for the assessment to follow. That condition covers, inter alia, smart solutions for file handling, which include various functions from indexing, searching, text and speech processing or linking data to other data sources, or AI systems used for translation of initial documents. In any case, AI systems used in high-risk use-cases listed in an annex to this Regulation should be considered to pose significant risks of harm to the health, safety or fundamental rights if the AI system implies profiling within the meaning of Article 4, point (4) of Regulation (EU) 2016/679 or Article 3, point (4) of Directive (EU) 2016/680 or Article 3, point (5) of Regulation (EU) 2018/1725. To ensure traceability and transparency, a provider who considers that an AI system is not high-risk on the basis of the conditions referred to above should draw up documentation of the assessment before that system is placed on the market or put into service and should provide that documentation to national competent authorities upon request. Such a provider should be obliged to register the AI system in the EU database established under this Regulation. With a view to providing further guidance for the practical implementation of the conditions under which the AI systems listed in an annex to this Regulation are, on an exceptional basis, non-high-risk, the Commission should, after consulting the Board, provide guidelines specifying that practical implementation, completed by a comprehensive list of practical examples of use cases of AI systems that are high-risk and use cases that are not. ### Recital 134 — deep fake transparency labelling obligations *Source: AI Act, aiact-rec-134-en, 2024-06-12 — https://overview.legal/posts/93950* Further to the technical solutions employed by the providers of the AI system, deployers who use an AI system to generate or manipulate image, audio or video content that appreciably resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful (deep fakes), should also clearly and distinguishably disclose that the content has been artificially created or manipulated by labelling the AI output accordingly and disclosing its artificial origin. Compliance with this transparency obligation should not be interpreted as indicating that the use of the AI system or its output impedes the right to freedom of expression and the right to freedom of the arts and sciences guaranteed in the Charter, in particular where the content is part of an evidently creative, satirical, artistic, fictional or analogous work or programme, subject to appropriate safeguards for the rights and freedoms of third parties. In those cases, the transparency obligation for deep fakes set out in this Regulation is limited to disclosure of the existence of such generated or manipulated content in an appropriate manner that does not hamper the display or enjoyment of the work, including its normal exploitation and use, while maintaining the utility and quality of the work. In addition, it is also appropriate to envisage a similar disclosure obligation in relation to AI-generated or manipulated text to the extent it is published with the purpose of informing the public on matters of public interest unless the AI-generated content has undergone a process of human review or editorial control and a natural or legal person holds editorial responsibility for the publication of the content. ## Guidance ### EDPB-EDPS Joint Opinion 5/2021 on the proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) *Source: EDPB, edpb-edps-joint-opinion-52021-on-the-proposal-for-a-regulation-of-the-en, 2021-06-18 — https://overview.legal/posts/126016 — original: https://www.edpb.europa.eu/documents/legislative-opinion/edpb-edps-joint-opinion-52021-on-the-proposal-for-a-regulation-of-the_en* 1 Adopted EDPB - EDPS Joint Opinion 5 /2021 on the proposal for a Regulation of the European Parliament and of the Council laying down harmo nised rules on artificial i ntelligence (Artificial Intelligence Act) 18 June 2021 2 Adopted Executive Summary On 2 1 April 2021, the European Commission presented its Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (hereinafter “the Proposal”) . The EDPB and the EDPS welcome… ### Guidelines 02/2021 on virtual voice assistants *Source: EDPB, edpb-guidelines-on-virtual-voice-assistants, 2021-07-07 — https://overview.legal/posts/38077 — original: https://www.edpb.europa.eu/documents/guideline/guidelines-022021-on-virtual-voice-assistants_en* A virtual voice assistant (VVA) is a service that understands voice commands and executes them or mediates with other IT systems if needed. VVAs are currently available on most smartphones and tablets, traditional computers, and, in the latest years, even standalone devices like smart speakers. VVAs act as interface between users and their computing devices and online services such as search engines or online shops. Due to their role, VVAs have access to a huge amount of personal... ### Guidelines 05/2022 on the use of facial recognition technology in the area of law enforcement *Source: EDPB, edpb-guidelines-on-the-use-of-facial-recognition technology-in-the-area-of-law-enforcement, 2023-05-17 — https://overview.legal/posts/38075 — original: https://www.edpb.europa.eu/documents/guideline/guidelines-052022-on-the-use-of-facial-recognition-technology-in-the-area-of_en* More and more law enforcement authorities (LEAs) apply or intend to apply facial recognition technology (FRT). It may be used to authenticate or to identify a person and can be applied on videos (e.g. CCTV) or photographs. It may be used for various purposes, including to search for persons in police watch lists or to monitor a person's movements in the public space. FRT is built on the processing of biometric data , therefore, it encompasses the processing of special categories ... ## Recent developments ### Swedbank refuses transparency in automatic interest calculation *Source: noyb - European Center for Digital Rights, 2025-02-27 — https://overview.legal/posts/53160 — original: https://noyb.eu/en/swedbank-refuses-transparency-automatic-interest-calculation* Data Subject Rights Nowadays, more and more banks set their interest rates automatically, and without any human intervention. But even the smallest inaccuracies can cost consumers thousands of additional euros. While EU law allows the use of such an automatic system in certain circumstances, companies must follow strict rules to protect people’s fundamental right to privacy. Banks, for example, would need to provide their customers with “meaningful information about the logic involved” in calcul ### De Autoriteit Persoonsgegevens publiceert een rapport over de risicoanalyse van de AVG (Algemene Verordening Gegevensbescherming). *Source: AEPD, 2022-10-11 — https://overview.legal/posts/51791* De GDPR-risicoanalyse is bedoeld om controllers en verwerkers te helpen bij het identificeren van de risicofactoren voor de rechten en vrijheden van de betrokkenen, wiens gegevens worden verwerkt. Het doel is om een eerste inschatting te maken van het inherente risico, inclusief de noodzaak om een Privacy Impact Assessment (DIA) uit te voeren, en om het resterende risico te schatten als maatregelen en beveiligingsmechanismen worden gebruikt om specifieke risicofactoren te verminderen. ## Literature ### Artificial Intelligence in Decision-making: A Test of Consistency between the “EU AI Act” and the “General Data Protection Regulation” *Source: Athens Journal of Law, 2025-01-02 — https://overview.legal/posts/132443 — original: https://doi.org/10.30958/ajl.11-1-3* The recent Regulation that sets down harmonised rules on Artificial Intelligence in the European Union, known as the "AI Act," includes a significant requirement for human oversight in high-risk AI systems during their use (art. 14). This requirement embodies the "human-in-command" approach, ensuring both legal and ethical compliance. The AI Act is intended to complement the General Data Protection Regulation (hereinafter GDPR), thereby forming a consistent and comprehensive legal framework. Thi ### General-Purpose AI under the EU AI Act: A Conceptual Allocation of Duties across the Value Chain *Source: SCRIPTed A Journal of Law Technology & Society, 2026-06-30 — https://overview.legal/posts/132370 — original: https://doi.org/10.2218/scrip.12300* This article examines how the final version of the EU Artificial Intelligence Act (“AI Act”, adopted 2024) allocates obligations across the AI value chain, with a focus on general-purpose AI (“GPAI”) or foundation models. It proposes a taxonomy of key actors – foundation model providers, fine-tuners, integrators, and deployers – and analyses the interfaces between them, including documentation tools (model cards, system cards) and logging requirements. Building on principles of control, foreseea ### HOW GDPR TREATS AUTOMATED DECISION-MAKING *Source: Journal Scientific and Applied Research, 2025-11-14 — https://overview.legal/posts/132599 — original: https://doi.org/10.46687/jsar.v28i1.435* This article examines how the General Data Protection Regulation (GDPR) regulates automated decision-making, including profiling, in the context of personal data processing. It analyzes the main provisions of Article 22 of the Regulation, as well as the conditions under which fully automated decisions that produce legal effects or significantly affect data subjects are permitted. The article highlights the rights of data subjects – the right to human intervention, the right to express their poin ### The Path of Formulating the Basic Law of Artificial Intelligence in China — Analysis of the Desirability of the EU Artificial Intelligence Act *Source: Studies in Law and Justice, 2023-09-01 — https://overview.legal/posts/132567 — original: https://doi.org/10.56397/slj.2023.09.09* The European Commission released the proposed Regulation on Artificial Intelligence (the EU AI Act) on 21 April 2021, which reflects the EU’s leadership orientation in establishing norms and standards in emerging fields, and also reflects the urgent need for legal unity of the EU as a unified market entity. The Act sets out harmonized rules for the development, placing on the market, and use of AI in the European Union. The ideas of a risk-based approach and experimental governance are of great ### REGULATION OF APPLIED ARTIFICIAL INTELLIGENCE IN BIOMEDICAL ENGINEERING AS A HIGH-RISK ARTIFICIAL INTELLIGENCE SYSTEM IN THE EU AI ACT *Source: AFMN Biomedicine, 2026-07-13 — https://overview.legal/posts/132435 — original: https://doi.org/10.65641/afmnai-2026-075* lt;p style= quot;text-align: justify; quot; gt; lt;span class= quot;a_GcMg font-feature-liga-off font-feature-clig-off font-feature-calt-off text-decoration-none text-strikethrough-none quot; gt;Artificial intelligence (AI) represents a global phenomenon changing all spheres of human life. Biomedical engineering is no exception, as many AI systems are applied to biomedical engineering inventions. The European Union has enacted the new EU AI Act, one of the world amp;rsquo;s first laws on AI. The ## Related topics - **Artificial Intelligence** — https://overview.legal/topics/ai AI systems and their implications for data protection - **Transparency** — https://overview.legal/topics/transparantie Openness about data processing activities - **AI Value Chain Actors and Roles** — https://overview.legal/topics/ai-value-chain-actors The content focuses on responsibilities distributed across different actors in the AI value chain. A dedicated topic for understanding the various actors, their - **Personal Data** — https://overview.legal/topics/persoonsgegevens Information relating to identified or identifiable natural persons - **Monitoring** — https://overview.legal/topics/monitoring Systematic observation and tracking of individuals - **Automated Decision-Making** — https://overview.legal/topics/geautomatiseerde-besluitvorming Processing involving automated decisions without human involvement --- Generated by overview.legal · https://overview.legal/topics/human-oversight-ai · 2026-08-22