[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85201-en":3,"doc-seo-85201-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},85201,962075114101,"Seraphina","https://ap-avatar.wpscdn.com/avatar/e000253a75eb197efd?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780044092746381165",8,"Research & Report","Intervenability as a Design Requirement for Autonomy and Oversight within Human-Centered AI","Intervenability is presented as a core design requirement for human-centered AI, enabling people to intervene and review AI behavior beyond options such as emergency shutdowns, workarounds, or system reconfiguration. The concept operationalizes controllability, autonomy, oversight, and keeping humans in the loop, supported by a taxonomy spanning from real-time control of automated processes to AI-based discrete, case-related decision-making. Interventions support development of human capabilities and guide technical improvement through reconfiguration, while enabling ongoing organizational learning and future AI that can perform interventions and transition smoothly to reconfiguration.","“Preprint of a book chapter in: Coursaris, C.K., Beringer, J., Léger, PM., Öz, B. (eds) The Design of Human-Centered Artificial Intelligence for the Workplace. Progress in IS. Springer, Cham (pp 143–166) . The final authenticated version is available at [https://doi.org/10.1007/978-3-031-83512-4_9](https://doi.org/10.1007/978-3-031-83512-4_9) ”  \nIntervenability as a Design Requirement for Autonomy  \nand Oversight within Human-Centered AI  \nThomas Herrmann [0000-0002-9270-4501]  \nInformation and Technology Management, IAW, Ruhr-Universität Bochum, Bochum,  \nGermany  \n[thomas.herrmann@rub.de](thomas.herrmann@rub.de)  \nAbstract. Based on the literature and several practical examples of possible AI  \napplications, we outline the concept of intervenability. This new phenomenon is  \nnot covered by emergency shutdowns, workarounds, or the reconfiguration of  \nautomated systems. Intervenability instantiates the principles of controllability,  \nautonomy, oversight, and keeping humans in the loop in the context of AI. We  \nprovide a taxonomy that encompasses a range of possibilities for intervening  \nactivities and differentiates them regarding the mental effort of the users. This  \ntaxonomy extends the scope of interventions from real-time control of automat  \ned processes to AI-based discrete case-related decision-making. This is in ac  \ncordance with human-centered AI, which seeks to combine human strengths  \nwith the usage of AI. We demonstrate how intervenability can potentially con  \ntribute to the ongoing development of human capabilities on the one hand and  \nto further technical improvement by reconfiguration of AI on the other. Explor  \ning and collaboratively reflecting on the effects of interventions as an integral  \npart of organizational practices is key to enabling this continuous improvement  \non both sides. Intervenability also provides further momentum for the design of  \nan AI that can help realize interventions on its own and advance a smooth tran  \nsition from intervention to reconfiguration of the AI.  \nKeywords: human-centered AI; autonomy; oversight; intervention, socio  \ntechnical design, human in the loop, organization in the loop  \n1 Introduction  \nA wide range of risks have been associated with the use of artificial intelligence. The  \nsuperannuation of jobs, the impairment of human autonomy and freedom of choice,  \nunintended damage from bad-quality AI, and bias in AI decisions that disadvantage  \ncertain social groups are prominent examples. The risks are particularly high when AI  \nis directly coupled with the physical environment, such as in mobility, manufacturing,  \nor weapons applications, or whenever decisions made by AI are implemented  \nautomatically without checks. One contribution to mitigating these risks is to provide opportunities for people to intervene when dealing with AI. DFKI 1 expert Paul Lukowicz said in an interview on the German newscast Tagesschau on June 14, 2023: ‘It is always problematic when AI does things not subject to human intervention or review.’  \nIn our opinion, the powers of review and intervention are inseparable, and reviewing is a precondition for continuous human control. One has to be able to react if one sees that something is going wrong. However, designing these possibilities for intervening is not trivial. Schmidt and Herrmann [1] propose intervention interfaces between humans on the one side and automated processes or automated decisionmaking on the other. Implementing intervention possibilities is quite complex, as we will explain with respect to intervention interfaces. It requires measures at the level of the AI itself and the technical infrastructure surrounding it, and at the level of organizational practice.  \nThe implementation of intervenability is thus a socio-technical challenge, and we will answer the questions of what kind of possible activities should be covered by intervenability and how intervention can be realized to be successful. Intervenabilit","cbCaii4raSewVv25","https://ap.wps.com/l/cbCaii4raSewVv25","pdf",816161,4,1,24,"English","en",105,"# Introduction\n## Risks of AI and the need for intervention\n## Intervention interfaces and socio-technical implementation\n## Controllability and human-in-the-loop oversight","[{\"question\":\"What is intervenability in the context of human-centered AI?\",\"answer\":\"Intervenability is the ability to intervene and review AI behavior in ways that extend beyond emergency shutdowns, workarounds, or reconfiguration. It operationalizes controllability, autonomy, oversight, and keeping humans in the loop for AI systems and decisions.\"},{\"question\":\"Why is intervenability considered a socio-technical challenge?\",\"answer\":\"Implementing intervenability requires measures at multiple levels: within the AI itself and surrounding technical infrastructure, as well as within organizational practice. The document treats this as a combined technical and organizational problem rather than a purely interface-level task.\"},{\"question\":\"How does the proposed taxonomy of interventions differ from real-time control?\",\"answer\":\"The taxonomy ranges from real-time control of automated processes to AI-based discrete, case-related decision-making. It also differentiates options by the mental effort required from users, addressing when monitoring and intervention should be necessary or exceptional.\"}]",1784201709,60,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"intervenability-as-a-design-requirement-for-autonomy-and-oversight-within-human-centered-ai","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":20},"https://docshare.wps.com/document/intervenability-as-a-design-requirement-for-autonomy-and-oversight-within-human-centered-ai/85201/",{"url":52,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-24","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is intervenability in the context of human-centered AI?","Question",{"text":75,"@type":76},"Intervenability is the ability to intervene and review AI behavior in ways that extend beyond emergency shutdowns, workarounds, or reconfiguration. It operationalizes controllability, autonomy, oversight, and keeping humans in the loop for AI systems and decisions.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why is intervenability considered a socio-technical challenge?",{"text":80,"@type":76},"Implementing intervenability requires measures at multiple levels: within the AI itself and surrounding technical infrastructure, as well as within organizational practice. The document treats this as a combined technical and organizational problem rather than a purely interface-level task.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the proposed taxonomy of interventions differ from real-time control?",{"text":84,"@type":76},"The taxonomy ranges from real-time control of automated processes to AI-based discrete, case-related decision-making. It also differentiates options by the mental effort required from users, addressing when monitoring and intervention should be necessary or exceptional.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,109,114,119,122,127,130,134],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":29,"slug":108},5,"Comic","comic",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]