AccScience Publishing / DP / Online First / DOI: 10.36922/DP026290034
Submit to DP
Cite this article
1
Download
45
Views
Supplementary Material
article_dp026290034_supplement_100c37.zip
Related Info Links
Google Scholar
More by Authors Links
Journal Browser
Volume | Year
Issue
Search
News and Announcements
View All
ARTICLE

From human oversight to effective control: A comparative documentary framework for high-impact AI systems

Karim Hardy1*
Show Less
1 Mines Paris - PSL, Center for Research on Risks and Crises (CRC), 1 rue Claude Daunesse, 06904 Sophia Antipolis , France
DP, 026290034 https://doi.org/10.36922/DP026290034
Received: 13 July 2026 | Revised: 21 September 2026 | Accepted: 22 September 2026 | Published online: 15 October 2026
© 2026 by the Author(s). This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution -Noncommercial 4.0 International License (CC-by the license) ( https://creativecommons.org/licenses/by-nc/4.0/ )
Download PDF
XML
Cite
Abstract

Human oversight is frequently required for high-impact artificial intelligence (AI), yet the presence of a human reviewer does not establish effective control. This study develops and applies the Oversight-to-Control (O2C) framework through a comparative documentary analysis of eight principal prescriptive instruments and nineteen public cases, represented by 41 verified oversight-control evidence statements. O2C separates six dependent functions—Observe, Detect, Interpret, Challenge and Decide, Intervene, and Recover and Learn—and links functional status to technical, cognitive, information/workflow, and organizational/institutional failure mechanisms. Interpret was below Present in all 18 assessable cases (nine Limited and nine Failed). Intervene was below Present in seventeen of nineteen cases, and Challenge and Decide in sixteen. The leading mechanisms were missing escalation or fallback (F08, ten cases), inadequate validation or monitoring (F12, nine cases), and organizational governance, incentive, or culture failure (F15, eight cases). Recover and Learn was Present in 12 of 18 assessable cases, showing that ex-post correction can coexist with weak real-time control. The contribution is a bounded diagnostic structure that connects information access, independent judgment, timely action, and recovery without treating procedural human involvement as evidence of control. The purposive corpus, public-record dependence, partial reproducibility of initial discovery, and single-researcher coding limit generalization and require independent, prospective validation.

Keywords
Human oversight
Effective control
High-impact AI
Algorithmic governance
Socio-technical systems
Funding
No funding was received for conducting this study or preparing this manuscript.
Conflict of interest
The author has no relevant financial or non-financial interests to disclose.
References
  1. Shneiderman B. Human-centered artificial intelligence: reliable, safe and trustworthy. Int J Hum Comput Interact. 2020;36(6):495-504. doi: 10.1080/10447318.2020.1741118
  2. Steen M, Timan T, Vethman S. We need better images of AI and better conversations about AI. AI Soc. 2025;40:3615-3626. doi: 10.1007/s00146-024-02101-z
  3. Green B. The flaws of policies requiring human oversight of government algorithms. Comput Law Secur Rev. 2022;45:105681. doi: 10.1016/j.clsr.2022.105681
  4. Laux J. Institutionalised distrust and human oversight of artificial intelligence: towards a democratic design of AI governance under the European Union AI Act. AI Soc. 2024;39:2853-2866. doi: 10.1007/s00146-023-01777-z
  5. Santoni de Sio F, van den Hoven J. Meaningful human control over autonomous systems: a philosophical account. Front Robot AI. 2018;5:15. doi: 10.3389/frobt.2018.00015
  6. Siebert LC, Lupetti ML, Aizenberg E, et al. Meaningful human control: actionable properties for AI system design. AI Ethics. 2023;3:241-255. doi: 10.1007/s43681-022-00167-3
  7. Romeo G, Conti D. Exploring automation bias in human-AI collaboration: a review and implications for explainable AI. AI Soc. 2026;41:259-278. doi: 10.1007/s00146-025-02422-7
  8. Alon-Barkat S, Busuioc M. Human-AI interactions in public sector decision making: automation bias and selective adherence to algorithmic advice. J Public Adm Res Theory. 2023;33(1):153-169. doi: 10.1093/jopart/muac007
  9. Schemmer M, Kühl N, Benz C, Bartos A, Satzger G. Appropriate reliance on AI advice: conceptualization and the effect of explanations. In: Proceedings of the 28th International Conference on Intelligent User Interfaces. Association for Computing Machinery; 2023:410-422. doi: 10.1145/3581641.3584066
  10. Miller T. Explanation in artificial intelligence: insights from the social sciences. Artif Intell. 2019;267:1-38. doi: 10.1016/j.artint.2018.07.007
  11. Murdoch WJ, Singh C, Kumbier K, Abbasi-Asl R, Yu B. Definitions, methods, and applications in interpretable machine learning. Proc Natl Acad Sci USA. 2019;116(44):22071-22080. doi: 10.1073/pnas.1900654116
  12. Hase P, Bansal M. Evaluating explainable AI: which algorithmic explanations help users predict model behavior? In: Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. Association for Computational Linguistics; 2020:5540-5552. doi: 10.18653/v1/2020.acl-main.491
  13. Poursabzi-Sangdeh F, Goldstein DG, Hofman JM, Wortman Vaughan J, Wallach H. Manipulating and measuring model interpretability. In: Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems. Association for Computing Machinery; 2021:237. doi: 10.1145/3411764.3445315
  14. Raji ID, Smart A, White RN, et al. Closing the AI accountability gap: defining an end-to-end framework for internal algorithmic auditing. In: Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. Association for Computing Machinery; 2020:33-44. doi: 10.1145/3351095.3372873
  15. Li B, Qi P, Liu B, et al. Trustworthy AI: from principles to practices. ACM Comput Surv. 2023;55(9):Article 177. doi: 10.1145/3555803
  16. Suresh H, Guttag JV. A framework for understanding sources of harm throughout the machine learning life cycle. In: Proceedings of the 1st ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization. Association for Computing Machinery; 2021:17. doi: 10.1145/3465416.3483305
  17. Bansal G, Wu T, Zhou J, et al. Does the whole exceed its parts? The effect of AI explanations on complementary team performance. In: Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems. Association for Computing Machinery; 2021:81. doi: 10.1145/3411764.3445717
  18. Amershi S, Weld D, Vorvoreanu M, et al. Guidelines for human-AI interaction. In: Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems. Association for Computing Machinery; 2019:3. doi: 10.1145/3290605.3300233
  19. Buçinca Z, Malaya MB, Gajos KZ. To trust or to think: cognitive forcing functions can reduce overreliance on AI in AI-assisted decision-making. Proc ACM Hum Comput Interact. 2021;5(CSCW1):188. doi: 10.1145/3449287
  20. Vasconcelos H, Jörke M, Grunde-McLaughlin M, Gerstenberg T, Bernstein M, Krishna R. Explanations can reduce overreliance on AI systems during decision-making. Proc ACM Hum Comput Interact. 2023;7(CSCW1):129. doi: 10.1145/3579605
  21. Jobin A, Ienca M, Vayena E. The global landscape of AI ethics guidelines. Nat Mach Intell. 2019;1(9):389-399. doi: 10.1038/s42256-019-0088-2
  22. Hagendorff T. The ethics of AI ethics: an evaluation of guidelines. Minds Mach. 2020;30(1):99-120. doi: 10.1007/s11023-020-09517-8
  23. Mittelstadt B. Principles alone cannot guarantee ethical AI. Nat Mach Intell. 2019;1(11):501-507. doi: 10.1038/s42256-019-0114-4
  24. Morley J, Floridi L, Kinsey L, Elhalal A. From what to how: an initial review of publicly available AI ethics tools, methods and research to translate principles into practices. Sci Eng Ethics. 2020;26(4):2141-2168. doi: 10.1007/s11948-019-00165-5
  25. Robbins S. What machines should not do. AI Soc. 2025;40:4093-4104. doi: 10.1007/s00146-024-02169-7
  26. Boshuijzen-van Burken C, de Vries M, Munyasya A. Autonomous military systems beyond human control: putting an empirical perspective on value trade-offs for autonomous systems design in the military. AI Soc. 2025;40:2507-2523. doi: 10.1007/s00146-024-02000-3
  27. European Parliament and Council of the European Union. Regulation (EU) 2024/1689 laying down harmonised rules on artificial intelligence. Off J Eur Union. Published July 12, 2024. Accessed June 24, 2026. https://eur-lex.europa.eu/eli/reg/2024/1689/oj/eng
  28. Tabassi E. Artificial Intelligence Risk Management Framework (AI RMF 1.0). National Institute of Standards and Technology (U.S.); 2023. doi: 10.6028/nist.ai.100-1
  29. National Institute of Standards and Technology. AI RMF Playbook. Accessed June 24, 2026. https://airc.nist.gov/airmf-resources/playbook/
  30. Organisation for Economic Co-operation and Development. OECD AI Principles. Updated May 2024. Accessed June 24, 2026. https://oecd.ai/en/ai-principles
  31. United Nations Educational, Scientific and Cultural Organization. Recommendation on the Ethics of Artificial Intelligence. Adopted November 23, 2021. Accessed June 24, 2026. https://unesdoc.unesco.org/ark:/48223/pf0000380455
  32. Council of Europe. Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law. CETS No. 225. Adopted May 17, 2024. Accessed June 24, 2026. https://www.coe.int/en/web/artificial-intelligence/the-framework-convention-on-artificial-intelligence
  33. Information Commissioner’s Office. Data protection audit framework: artificial intelligence¡ªhuman review. Accessed June 24, 2026. https://ico.org.uk/for-organisations/advice-and-services/audits/data-protection-audit-framework/toolkits/artificial-intelligence/human-review/
  34. European Commission High-Level Expert Group on Artificial Intelligence. Ethics Guidelines for Trustworthy AI. European Commission; 2019. https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai
  35. Wisconsin Supreme Court. State v Loomis. 2016 WI 68;881 NW2d 749.
  36. New York State Department of Financial Services. Report on Apple Card Investigation. Published March 2021. Accessed June 24, 2026. https://www.dfs.ny.gov/reports_and_publications/202103_report_apple_card_investigation
  37. Beede E, Baylor E, Hersch F, et al. A human-centered evaluation of a deep learning system deployed in clinics for the detection of diabetic retinopathy. In: Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems. Association for Computing Machinery; 2020:1-12. doi: 10.1145/3313831.3376718
  38. Wong A, Otles E, Donnelly JP, et al. External validation of a widely implemented proprietary sepsis prediction model in hospitalized patients. JAMA Intern Med. 2021;181(8):1065-1070. doi: 10.1001/jamainternmed.2021.2626
  39. Obermeyer Z, Powers B, Vogeli C, Mullainathan S. Dissecting racial bias in an algorithm used to manage the health of populations. Science. 2019;366(6464):447-453. doi: 10.1126/science.aax2342
  40. Zhang Y, Liao QV, Bellamy RKE. Effect of confidence and explanation on accuracy and trust calibration in AI-assisted decision making. In: Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. Association for Computing Machinery; 2020:295-305. doi: 10.1145/3351095.3372852
  41. Lai V, Chen C, Smith-Renner A, Liao QV, Tan C. Towards a science of human-AI decision making: an overview of design space in empirical human-subject studies. In: Proceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency. Association for Computing Machinery; 2023:1369-1385. doi: 10.1145/3593013.3594087
Next article in this issue
Share
Back to top
Design+, Electronic ISSN: 3060-8953 Published by AccScience Publishing
We use cookies on our website to ensure you get the best experience.
Read more about our cookies here
Accept