Ethical Design Futures Framework

• Ethical Design Futures Framework is a conceptual orientation that embeds ethics across the design, adoption, and evolution of sociotechnical systems.
• It employs multi-scale ethical reasoning and diverse methodological forms, including participatory tools and lifecycle-wide evaluations.
• The framework is applied in domains like generative AI, IoT, and medical imaging, promoting proactive innovation and responsible governance.

The Ethical Design Futures Framework is a conceptual orientation for embedding ethics into the design, adoption, adaptation, and long-term shaping of sociotechnical systems rather than treating ethics as a compliance layer added after technical development. In the literature surveyed here, it appears as a family of closely related approaches that connect ethical reflection to design practice, organizational context, stakeholder participation, and future-oriented judgment. One major strand frames the problem through generative AI in creative production; another through augmentation technologies and AI in technical and professional communication; others extend the same orientation to personal data management, embedded community research, IoT, medical imaging, adaptive XR, and speculative HRI (Hofman, 2024, Duin et al., 13 Aug 2025, Toussaint et al., 2021).

1. Emergence and scope

A central move in this body of work is the shift from developer-centered AI ethics to ethics as a user-, process-, and practice-centered activity. Geert Hofman’s account of generative AI in creative production explicitly argues that technology consumers in design, marketing, communication, and product development need a practical ethical framework that fits real workflows rather than abstract principle lists, especially at the “consumption stage” of the AI lifecycle where downstream users inherit an “upstream legacy” of governance and ethical issues from model developers and platform providers (Hofman, 2024). In parallel, the chapter on augmentation technologies and AI defines an Ethical Design Futures Framework as a guide for reframing professional practice and pedagogy to “promote digital and AI literacy surrounding the ethical design, adoption, and adaptation of augmentation technologies,” with a particular focus on technical and professional communication and on technologies being “adapted and normalized for everyday users” (Duin et al., 13 Aug 2025).

This literature also broadens the scale at which ethics is understood. “Towards an Ethical Framework in the Complex Digital Era” argues that ethical frameworks based only on the individual level are no longer sufficient and proposes a multi-scale ethical vision linking individuals, communities, and global systems through “protective principles,” “actionable principles,” and “projection principles” (Pastor-Escuredo et al., 2020). “Peril v. Promise: IoT and the Ethical Imaginaries” similarly criticizes both top-down ethical frameworks and solutionist approaches that treat ethical issues as technical problems, arguing instead for attention to situated capabilities, distributed responsibility, and ethical imaginaries that shape how futures are understood before they fully materialize (Ustek-Spilda et al., 2019).

Taken together, these contributions frame the Ethical Design Futures Framework less as a single canonical model than as an evolving design-ethics field organized around situated practice, ethical anticipation, and explicit engagement with future consequences. This suggests a framework class in which ethics is inseparable from design activity, stakeholder relations, and the social trajectories opened or closed by technological systems (Öz et al., 6 Jun 2025).

2. Core principles and scales

Across the literature, several recurring commitments define the framework’s normative center. Hofman’s formulation is explicitly anchored in responsibility, anticipation, reflection, and playful exploration, with ethics embedded “across the human creative process and the AI lifecycle” rather than treated as compliance (Hofman, 2024). The augmentation-technologies chapter centers “ethical design, adoption, and adaptation,” “human-centered, humane futures,” “digital and AI literacy,” and attention to “human subjects, contexts, and rhetorical strategies proposed for them by external actors” (Duin et al., 13 Aug 2025). The complex-digital-era paper adds sustainability, resilience, multiculturality, multi-level society, and truth-grounded action as long-range ethical commitments (Pastor-Escuredo et al., 2020).

Scale is equally central. Hofman’s “compass” explicitly organizes ethical reflection across micro, meso, and macro levels: effects on the team and immediate practices, effects on clients and partners, and societal and environmental impacts (Hofman, 2024). The lifecycle framework for AI in medical imaging distributes ethical inquiry across data collection, data processing, model training, model evaluation, and deployment, while repeatedly foregrounding privacy, security, fairness, transparency, explainability, accountability, autonomy, accessibility, clinical safety, regulatory compliance, and ongoing monitoring (Khan et al., 6 Jul 2025). The responsible-foresight position paper generalizes this multi-scale stance through sustainability, equity and intergenerational justice, systems thinking, adaptability and responsiveness, exploration of multiple futures, continuous monitoring and feedback loops, scientific rigor, and data integrity (Perez-Ortiz, 26 Nov 2025).

A further common element is plural ethical reasoning rather than reduction to a single norm. Hofman operationalizes six ethical theories as lenses—virtue ethics, deontology, consequentialism or utilitarianism, contract ethics, care ethics, and existential ethics—so that teams can surface multiple ethical dimensions rather than derive one final answer (Hofman, 2024). The community-led feminist ethics paper reaches a related conclusion from another direction by grounding ethics in situated knowledges, standpoint theory, intersectionality, participatory methods, and care ethics, and by describing ethics as a “process of frameworking — rather than a fixed framework” (Henriques et al., 2024).

3. Process architectures and methodological forms

The framework is most fully elaborated when ethics is tied to concrete process architectures. Hofman’s model combines the Double Diamond (“Discover, Define, Develop, Deliver”), an AI lifecycle perspective (“creation, production, implementation, instruction, consumption, and policymaking”), six ethical lenses, a “compass” based on “zooming in and zooming out,” and three modes of action: “Use, think, try” (Hofman, 2024). In this design, zooming out occurs at “big transition moments” such as the start of the process, the transition between problem and solution phase, each iteration, and the end of the process; zooming in occurs before and after phases where AI tools are used, during planning, in experiments, and whenever a concern is detected (Hofman, 2024).

The augmentation-technologies chapter offers a different but complementary architecture organized into three major sections: “Understand (rhetorics of) augmentation technologies,” “Build literacies,” and “Design ethical futures.” Its central purpose is to “reframe professional practice and pedagogy to promote digital and AI literacy surrounding the ethical design, adoption, and adaptation of augmentation technologies,” and it supplements this with taxonomic distinctions between cognitive, sensory, physical, and emotional augmentation, as well as on-body, in-body, and around-body relations (Duin et al., 13 Aug 2025). The paper on responsible computational foresight adds a policy-oriented methodological layer in which forecasting, prediction markets, world simulation, simulation intelligence, scenario building, participatory futures, futures literacy, and hybrid intelligence are combined to move “from prediction to foresight” (Perez-Ortiz, 26 Nov 2025).

Several works also contribute process-specific subframeworks. In medical imaging, the five-stage lifecycle model links ethical safeguards to specific development phases and asks structured quality-check questions at each phase (Khan et al., 6 Jul 2025). In cross-disciplinary algorithm ethics, the “Four C’s Framework” identifies computability, complexity, consistency, and controllability as feasibility conditions for ethical algorithmic governance, arguing that ethical requirements must be formalizable, scalable, coherent, and steerable if they are to function beyond principle statements (Perrier, 2021). Across these models, ethics appears not as a single checkpoint but as a recursive structure of inquiry, iteration, and governance.

4. Participatory and pedagogical instruments

A distinctive feature of the Ethical Design Futures Framework literature is the use of artifacts, workshops, and training devices that make abstract ethical issues discussable. “Design Considerations for Data Daemons” shows how ethical personal data management can be co-created through design futures, the Future workshop method, design fictions, and a fictional mediating construct called Data Daemons. The paper distills three design considerations—“Supported personal choice and control,” “Interaction and collaboration through negotiation,” and “Accountability and adaptation to value change”—from an iterative loop linking concept, fiction, and design reflection (Toussaint et al., 2021). Its contribution is methodological: ethical abstraction becomes participatory design material.

“The Moral-IT Deck” performs a similar translation for technologists through a 52-card deck organized into Security, Ethics, Privacy, and Law, paired with an impact-assessment-board process structured around identifying risks, assessing significance, establishing safeguards, and exploring implementation challenges. The study reports that the cards support “ethical clustering,” “ethical sorting and comparison,” and provide “appropriate anchors” for discussion while revealing the intertwined nature of ethics in design (Urquhart et al., 2020). “Navigating Ethics and Power Dynamics through Participant-Designer Journey Mapping” offers the Participant-Designer Journey Map, built around an Alignment phase plus the Double Diamond and structured through “Understanding,” “Mapping,” and “Reflecting” sections to expose value alignment, hidden power dynamics, direct interaction moments, power redistribution, collaborative decision-making, diversity and inclusion, and moments of tension (Tejo et al., 2024).

Pedagogical applications extend this participatory orientation. The adaptive-XR project bRight-XR uses Design-Based Research to build a heuristic evaluation matrix and self-assessment toolkit for designers, with matrix dimensions including disciplinary fields, type of interaction, modalities of measurement, temporality of effect, reliability indicator, and robustness of usage, all developed “considering the ethics of care” (Rouyer et al., 2024). The augmentation-technologies chapter likewise positions educators, students, and practitioners as explicit targets of an Ethical Design Futures Framework, while the review of design ethics in practice emphasizes that ethics is “cultivable” through collaborative methods, reflective routines, and organizational support (Duin et al., 13 Aug 2025, Öz et al., 6 Jun 2025).

5. Application domains and recurrent ethical tensions

The framework has been articulated across a wide range of domains, but the ethical tensions are strikingly recurrent. In generative AI for creative production, concerns include “machines replacing human creativity,” over-reliance and deskilling, bias, unfairness, opacity, misinformation, hallucinations, privacy, security, intellectual property, originality, sustainability, and the tension between ethical rigor and creative flow (Hofman, 2024). In augmentation technologies, the same orientation appears through concerns about algorithmic bias, discrimination, surveillance, data extraction, privacy violations, digital divides, opaque AI decision-making, and the rhetoric that normalizes enhancement technologies before their consequences are understood (Duin et al., 13 Aug 2025).

Other domains expose additional layers. IoT ethics is described through “peril” and “promise” imaginaries that defer responsibility while narrowing ethical attention to privacy and security and neglecting equity, equality, trustability, and broader social and economic issues (Ustek-Spilda et al., 2019). The soft robotic wearable study “Sumbrella” surfaces exploitation, surveillance, privacy, consent, commercialization of intimate data, autonomy loss, stigma, public acceptability, and identity-sensitive risks in public space, especially where concealment and visibility are interpreted differently across gender and ethnic identities (Ingold et al., 29 Dec 2025). In medical imaging, the central issues are patient consent, anonymization, access control, demographic underrepresentation, model bias, stakeholder legibility, and post-deployment oversight (Khan et al., 6 Jul 2025).

A recurrent controversy concerns whether ethics should be framed mainly as harm reduction. The hope-centered workshop paper argues that design should move beyond purely reactive problem solving and “harm mitigation” toward “deliberate hope,” using “goal-directed,” “pathways,” and “agentic” thinking to cultivate proactive futures without collapsing into “unquestioned optimism or naive trust” (Kim et al., 10 Mar 2025). The practice-turn review reaches a related conclusion in HCI by characterizing ethics as inherent, contextually dependent, evolving, and cultivable rather than static, universalized, and external to practice (Öz et al., 6 Jun 2025).

6. Debates, limitations, and future directions

Despite its breadth, the Ethical Design Futures Framework remains conceptually unsettled and often exploratory. Hofman’s framework is explicitly conceptual and not empirically validated; its transparency label is more concrete than its sustainability label, and its questions are illustrative rather than exhaustive (Hofman, 2024). The augmentation-technologies chapter is also an introductory conceptual chapter rather than a fully operational method, and many details are deferred to later chapters (Duin et al., 13 Aug 2025). The hope workshop is a proposal rather than a completed empirical study, and the embedded-research workshop paper is agenda-setting rather than a validated framework (Kim et al., 10 Mar 2025, Carter et al., 2024).

The literature also identifies persistent structural questions. The community-led feminist ethics paper criticizes static, researcher-centered, and decontextualized ethics and calls instead for a “multidimensional adaptive model for ethics in HCI design,” but it deliberately resists closure in favor of “frameworking” (Henriques et al., 2024). The IoT paper warns that ethics can be postponed to future regulation, future technological maturity, or conscious consumers, thereby shifting responsibility rather than resolving it (Ustek-Spilda et al., 2019). The review of ethics in practice calls for concept-building empirical research, longitudinal studies, stronger contextualization, more attention to entanglements of human and nonhuman actors, specialized design methods for cultivating ethics, and ways of making ethics engaging rather than purely burdensome (Öz et al., 6 Jun 2025).

Future directions in this field therefore tend to converge on a common agenda: stronger empirical validation; richer participatory and community-led governance; lifecycle-wide accountability; methods for anticipation, reflection, and post-deployment monitoring; and better integration of foresight, scenario work, and stakeholder capability-building. A plausible implication is that the Ethical Design Futures Framework will continue to develop not as a single universal standard, but as a layered and interoperable set of process models, participatory devices, and governance practices for designing ethical futures under uncertainty (Perez-Ortiz, 26 Nov 2025).