Civic AI and the 6-Pack of Care: Reimagining AI alignment for relational health
In recent years, the field of AI alignment has developed rapidly. Much of this progress has been rooted in utilitarian frameworks and hierarchical models of control, approaches that prioritise optimising outcomes and ensuring that artificial agents remain safely directed by human objectives.
These methods remain indispensable, particularly for developing technical safeguards. Yet, as our technologies, societies, and people's lives become increasingly interconnected, it is clear that no single framework will be enough. AI does not operate in isolation: it is developed and exists within diverse communities, subject to systems of power and a plurality of values. For this reason, relational and process-based approaches offer an important complement, enriching alignment efforts by foregrounding interdependence, interconnectedness, dialogue, and mutual responsiveness.
At the Institute for Ethics in AI’s Accelerator Fellowship Programme, we are working on just such an approach, which we call ‘Civic AI’. It draws on Joan Tronto’s influential work in the ethics of care alongside Ambassador Audrey Tang’s ⿻ Plurality vision, which emphasises collaborative diversity. Together, these perspectives suggest that AI ethics should not only prevent harm or optimize for efficiency, but also cultivate ‘civic care’, embedding values of attentiveness, solidarity, and relational health directly into the design and governance of intelligent systems and fostering democratic participation with AI.
At its foundation lies the recognition of human interdependence and the conviction that wellbeing and dignity are not individual luxuries, but shared goods. Civic AI extends care beyond person-to-person interactions to encompass relationships between humans and AI systems, as well as among AI systems themselves.
Importantly, this framework does not seek to displace existing alignment paradigms. Rather, it provides additional conceptual and practical tools, building on real-world precedents such as the vTaiwan platform for participatory democracy, and echoing calls from Cooperative AI researchers for scalable and inclusive governance structures.
The 6-Pack of Care
At the heart of this approach is what we call the 6-Pack of Care, six principles that connect Tronto’s care ethics to AI alignment and reframe the latter as an ongoing, relational process. Each principle is designed to address the challenge of horizontal coordination, where multiple agents with differing perspectives must coexist, cooperate, and adapt.
Pack 1: Attentiveness in Recognition
The first step in care is noticing. For AI, this means cultivating attentiveness to the needs and contexts of diverse stakeholders. In practice, this could involve building sensemaking tools that reduce information asymmetries among agents, ensuring that systems amplify, not marginalise, voices across interconnected networks.
Pack 2: Responsibility in Engagement
Caring also requires taking responsibility. In multi-agent environments, AI systems must be able to accommodate and be subject to forms of accountability, complementing existing mechanisms of trust and credible commitments. Responsibility here is less about rigid control and more about sustained engagement with dynamic relational contexts.
Pack 3: Competence in Action
Good intentions are insufficient without competence. For AI, competence entails context-sensitive interventions that are both feasible and effective. This includes designing tools that promote cooperation while discouraging collusion, thereby strengthening democratic processes in collective decision-making.
Pack 4: Responsiveness in Adaptation
Care involves listening and adapting. AI systems must be responsive to feedback, willing to recalibrate their behaviours in the light of community input. This responsiveness mirrors Tronto’s 'care-receiving' phase, enabling AI to evolve symbiotically rather than rigidly persisting in predetermined trajectories. In ecological terms, this resembles a local kami (referring to concept from the Shinto religion: a spirit quietly residing and watching over a specific patch of land), a humble yet adaptive presence within a polycentric system.
Pack 5: Solidarity in Community
Care is not only dyadic: it is collective. Civic AI calls for solidarity, expressed through respect, communication and the pursuit of shared flourishing. This principle operationalizes plurality at the infrastructural level, ensuring that normative systems can transform potential conflict into resilient, inclusive collaboration.
Pack 6: Symbiosis in Horizon
The final principle envisions AI as a shared good, created “of, by, and for” communities. This entails embedding logics of sufficiency, what we call 'enoughness', and resisting extractive tendencies. By prioritising long-term symbiosis, AI can become part of a decentralised democratic defence, contributing to futures where civic care is not incidental but foundational.
Toward Relational Health
These six principles can be imagined as core muscles in a civic “6-pack”: just as physical endurance requires balanced training, relational endurance requires cultivating each form of care in concert. Together, they provide the strength necessary for AI systems to support not only technical alignment, but also what we might call relational health: the ongoing maintenance and flourishing of human-human and human-AI relationships in pluralistic societies.
A Call for Collaboration
We see Civic AI not as a finished framework, but as an invitation. By integrating the ethics of care into design, governance, and practice, scholars, practitioners and civil society alike can help shift the trajectory of AI toward fostering solidarity and dignity. This approach stands alongside, rather than against, other ethical traditions, enriching a plural conversation about what it means to align AI with human values.
In short, alignment is not only a technical problem to be solved: it is also a civic project to be lived. By embedding care into our AI systems, we may move closer to a future in which technology sustains our shared humanity rather than undermining it.
Useful links
The project is further outlined on our 6-Pack of Care website: 6pack.care
You can also now listen to our podcast announcing the 6-Pack of Care, featuring Ambassador Tang and Dr Green: 6-Pack of Care podcast
You can find out more about Ambassador Audrey Tang on our website.
We are really pleased to announce our first ‘AI and Ethics of Care Conference’ to be held on the 25 March 2026, at Oxford University. More information will be available on our Events page soon.
Notes for Editors
Ambassador Audrey Tang and Director of Research Caroline Green are available for interviews. For more information, please contact: aiethicscomms@philosophy.ox.ac.uk
Find out more: https://www.oxford-aiethics.ox.ac.uk/
About the University of Oxford
Oxford University has been placed number 1 in the Times Higher Education World University Rankings for the ninth year running, and number 3 in the QS World Rankings 2024. At the heart of this success are the twin-pillars of our ground-breaking research and innovation and our distinctive educational offer.
Oxford is world-famous for research and teaching excellence and home to some of the most talented people from across the globe. Our work helps the lives of millions, solving real-world problems through a huge network of partnerships and collaborations. The breadth and interdisciplinary nature of our research alongside our personalised approach to teaching sparks imaginative and inventive insights and solutions.
Through its research commercialisation arm, Oxford University Innovation, Oxford is the highest university patent filer in the UK and is ranked first in the UK for university spinouts, having created more than 300 new companies since 1988. Over a third of these companies have been created in the past five years. The university is a catalyst for prosperity in Oxfordshire and the United Kingdom, contributing £15.7 billion to the UK economy in 2018/19, and supports more than 28,000 full time jobs.
