Behind the Paper: Why I Wrote From Assistants to Agents
Published in Philosophy & Religion
Explore the Research
From assistants to agents: a relational framework for human–AI co-agency
When we talk about artificial intelligence, much of the debate still oscillates between two familiar images. On one side, AI is treated as a sophisticated tool: something that supports human action but remains external to responsibility. On the other side, AI is sometimes described in near-anthropomorphic terms, as if it were becoming an autonomous actor in its own right. Both views, in different ways, seemed incomplete to me.
What increasingly interested me was a more difficult question: what happens when AI is neither merely assistive nor fully autonomous, but instead operates within structured systems of delegation, supervision, and institutional responsibility? In other words, what happens when action is distributed across humans, artificial systems, and governance structures?
That question became the starting point for my article, From Assistants to Agents: A Relational Framework for Human–AI Co-Agency, now published in AI and Ethics.
The central intuition behind the article is simple: the rise of agentic AI requires us to rethink agency relationally. In many real-world settings, AI does not act in isolation. Its apparent initiative is enabled, bounded, and interpreted through human decisions, institutional goals, technical design, and governance mechanisms. This means that responsibility cannot be understood by looking only at the machine, or only at the individual user. It has to be understood through the structure of the broader sociotechnical system.
To make that argument more operational, I proposed a four-dimensional framework built around initiative, decision scope, oversight, and responsibility attribution. These dimensions are intended to help clarify where delegation begins, how far it extends, how supervision is preserved, and how accountability should remain anchored even when systems take on greater initiative.
One reason I felt this mattered is that discussions about AI governance often remain either too abstract or too procedural. We speak in large terms about ethics, trust, regulation, and policy, but we often lack a practical conceptual bridge between theory and institutional design. I wanted this work to contribute to that middle layer: not only asking what AI is, but how human–AI arrangements should be evaluated when decisions, actions, and consequences are distributed.
I was also concerned that the vocabulary of “autonomy” can sometimes obscure more than it reveals. The issue is not simply whether a system appears autonomous. The more important question is whether institutions remain capable of structuring delegation responsibly. A system may appear highly capable while still depending on weak governance. Conversely, human accountability can be eroded not because humans disappear, but because roles, boundaries, and responsibility structures become blurred.
In that sense, the challenge posed by agentic AI is not machine agency in isolation. It is the alignment of delegated action with meaningful human oversight and institutional responsibility. More broadly, I hope this work contributes to a more grounded and rigorous conversation about how societies, organizations, and governance systems should respond as AI becomes more agentic.
Read the article
SharedIt: https://rdcu.be/fgHrc
DOI: https://doi.org/10.1007/s43681-026-01111-5
Independent researcher focused on AI ethics, philosophy of technology, governance, and sociotechnical systems. I am particularly interested in human–AI co-agency, accountability, and how responsible AI can be made meaningful in real institutional environments for the benefit of humanity.
Follow the Topic
-
AI and Ethics
This journal seeks to promote informed debate and discussion of the ethical, regulatory, and policy implications that arise from the development of AI. It focuses on how AI techniques, tools, and technologies are developing, including consideration of where these developments may lead in the future.
Related Collections
With Collections, you can get published faster and increase your visibility.
Participatory AI: Co-Designing Sociotechnical Systems
AI systems have become pervasive and deeply integrated into the fabric of social life. As they influence crucial everyday activities and decision-making processes, they hold the power to both support and harm people. Recognizing their dual potential and their sociotechnical nature is essential to rethinking how such systems are designed and how relevant stakeholders interact with them, fostering responsible and trustworthy human–AI interactions.
This topical collection examines how participatory approaches might address risks and limitations in AI-powered technologies by engaging diverse stakeholders in the design process. The collection explores what methods Participatory AI offers for shaping systems that better align with human values and community principles, while critically examining the challenges and tensions that these approaches encounter in practice.
Aim and scope
Drawing on the sociotechnical tradition that conceives social and technical elements as co-constructed, the aim of the collection is to bring together works at the intersection of sociotechnical studies and participatory design, exploring how AI and digital systems can be co-designed to reflect shared values, accountability, and agency. In particular, we welcome contributions that explore how participatory approaches can make AI systems more aligned with, responding to, and driven by specific users, communities, and contexts, rather than pursuing universal solutions. This collection emphasizes the inclusion of stakeholders in the earliest stages of decision-making, including discussions on whether a technology should be developed in the first place—welcoming submissions that ask whether these technologies are truly needed or wanted by communities.
Participatory AI is conceived not as a binary label but as a spectrum of practices and degrees of involvement, encompassing a variety of methods, intensities, and moments of engagement. Participatory approaches provide ways not only to design with and for communities to prevent bias from the outset, but also to perform bias control and mitigation in already deployed or existing AI systems, as well as possibilities of designing by communities. To advance these methods, we are also interested in critical discussions of the tensions and challenges they encounter in practice, such as the resource-intensive nature of genuine participation, how to address power imbalances that persist even in participatory settings, and how to mediate between community and individual values.
Areas of Interest
We welcome technical and non-technical submissions with theoretical, methodological, or experimental contributions, explicitly encouraging interdisciplinary submissions.
Topics of interest include:
- Methods, frameworks, and design solutions for participatory AI (co)design
- Experiments, simulations, prototypes, or case studies of co-design processes in AI development
- Strategies for balancing individual and collective needs in AI design
- Critical reflections on the motivations, challenges, and limitations of participatory approaches
- Analyses of power dynamics and ethical considerations in participatory AI design
- Experiences and lessons learned from co-design and stakeholder engagement
- Assessing AI impacts on diverse stakeholders through participatory approaches and stakeholder engagement
This topical collection is based on the previously organized workshop Mind the AI GAP: Co-designing sociotechnical systems (https://aigap2025.isti.cnr.it/) hosted at the 4th International Conference on Hybrid Human-Artificial Intelligence, 2025, Pisa (Italy) but is also open to other non-listed topics closely aligned with the overall scope of the collection.
Publishing Model: Hybrid
Deadline: Jun 30, 2026
AI Ethics for Children and Adolescents
This topical collection invites contributions that critically examine how central concepts and theories of AI ethics function when applied to children and adolescents, and where their limits become visible. While terms such as trust, explainability, informed consent, privacy, bias, justice, and well-being are well established in AI ethics, they are usually developed with adult users and decision-makers in view, which means that in contexts concerning children and adolescents they frequently rest on assumptions that do not hold or at least require critical examination.
Children and adolescents encounter AI systems under conditions of developing autonomy, heightened vulnerability, and dependence on others, which does not mean, however, that they are merely passive objects of protection – rather, they possess emerging forms of agency and a moral right to participation and development. Ethical analysis must therefore go beyond simple transfers of adult-centered frameworks and instead ask how AI ethics concepts must be specified, adapted, or fundamentally reconceived in developmentally appropriate and relational ways, whereby it is likely to emerge that such adaptations are not only relevant for children and adolescents but can also enrich the general debate.
We welcome submissions engaging in conceptual and normative analysis, as well as ethically informed empirical work. Contributions may focus on individual concepts, compare different ethical approaches, or explore concrete application contexts, with particular welcome given to work that makes explicit which assumptions about agency, competence, responsibility, or rationality are embedded in existing AI ethics frameworks and how these assumptions are challenged by childhood and adolescence. Also of interest are contributions addressing the question of how AI systems must be designed to meet the particular needs and rights of children and adolescents, or examining what governance structures are required to ensure child-sensitive AI.
Topics
Topics may include, but are not limited to:
• Trust and trustworthiness of AI systems in childhood and adolescence, including questions of overtrust, emotional attachment, and manipulative design strategies
• Explainability and transparency under conditions of developing cognitive capacities, whereby the danger of "explainability washing" must also be considered
• (Informed) consent, shared decision-making, and participation, including the question of how concepts such as transitional paternalism are to be evaluated ethically
• Privacy, surveillance, and data protection for children and adolescents, particularly in the context of digital phenotyping and other data-intensive applications
• Bias, discrimination, and justice affecting marginalized children, whereby intersectional perspectives should also be taken into account
• AI and the well-being of children and adolescents, including the question of socialization effects of AI
• Autonomy development, vulnerability, and dependence in AI-mediated environments, whereby the role of human relationships in an AI-permeated childhood must also be reflected upon
• Ethical governance and child-sensitive AI design, including the question of democratic participation of children and adolescents in decisions about their technological future.
Please find a detailed call for papers and submission guidelines at https://link.springer.com/journal/43681/updates/27841622.
Publishing Model: Hybrid
Deadline: Nov 30, 2026
Please sign in or register for FREE
If you are a registered user on Research Communities by Springer Nature, please sign in