Daejeon, July 27–31
As vast portions of online content disappear due to technical, financial, and organizational factors, web archiving has become a proactive strategy against ‘digital decay’, sustaining public knowledge and cultural memory. Web archives now support evidentiary, journalistic, advocacy, research or other purposes. Yet the rapid expansion of AI-driven crawling, large-scale capture, multimodal scraping, and platform-based restrictions raises complex questions about participation, responsibility, and justice in digital memory practice.
Existing discussions of web archiving ethics often treat privacy, copyright, or collection bias as discrete issues. However, ethical challenges in web archiving rarely exist in isolation. Instead, they arise from layered interactions among archived materials, diverse stakeholders (e.g., creators, users, archivers, platforms), and broader sociotechnical environments shaped by AI systems. Without a systematic structure for understanding these interactions, ethical debates remain fragmented, limiting the field’s capacity to diagnose problems and design governance mechanisms that reflect the realities of digital culture.
To address this gap, this study develops a two-dimensional analytical framework for archival ethics and applies it to web archiving. The framework is grounded in four ethical theories—virtue ethics, deontological ethics, utilitarianism, and care ethics, and is derived from qualitative analysis of seven major archival ethics policies worldwide. Together, these foundations support a conceptual structure that organizes ethical concerns along two dimensions:
The object-oriented dimension, which focuses on the values that archival practice should uphold in relation to its external objects—records, stakeholders, and society;
The self-oriented dimension, which emphasizes the virtues, professional capacities, and action strategies required of archivers to realize those values in practice.
The following extended abstract outlines the construction of this framework and demonstrates how it enhances the identification of ethical issues in web archiving while guiding governance strategies suitable for participatory, AI-mediated digital memory work.
As an important component of digital memory practice, web archiving has raised a wide range of ethical concerns. Scholars from multiple disciplines have examined issues related to user privacy and informed consent, particularly in relation to user-generated content and personal data on social media platforms. Others have highlighted copyright tensions, biased representation, and gaps arising from selective or uneven crawling. Additional concerns focus on how web archiving may inadvertently reproduce social inequalities, excluding marginalized voices or reinforcing existing power structures.
While these contributions are valuable, the field still lacks a unifying framework capable of integrating diverse ethical questions and situating them within broader social and technological contexts. Issues such as informed consent and the misrepresentation of social inequalities in archived web content both fall within the ethical realm, yet they involve fundamentally different concerns—the former centers on respecting and protecting individual rights, whereas the latter examines how web archiving shapes society and reinforce or challenge existing power structures. Ethical analysis, therefore, requires multiple perspectives—rights, obligations, societal impact, participant relations, and the evolving influence of AI—along with clarity about how these perspectives interrelate.
To construct an analytical basis for ethical reasoning, this study draws from four ethical theories:
Virtue ethics, which emphasizes the moral character, integrity, and cultivated judgment of archivers;
Deontological ethics, which centers on duties, rights, and principled constraints in archival action;
Utilitarianism, which focuses on consequences and societal benefits, particularly in collective memory and public good;
Care ethics, which foregrounds relational responsibilities and attention to contextual vulnerability.
Collectively, they give rise to two overarching ethical dimensions of archival practice. The first begins with the actor and concerns the self, focusing on the virtues archivers should cultivate and the professional capacities they should possess to enact them. The second begins with the objects or entities with which the actor interacts, emphasizing a relational and responsibility-oriented perspective that considers the duties and obligations arising from these interactions. These two dimensions are not mutually exclusive but conceptually complementary.
Based on the two dimensions, we analyzed seven influential archival ethics policies from international and national associations across Europe, North America, and Oceania. Using NVivo 12 software, we conducted qualitative coding that involved: reading the policies, tagging key phrases and text segments, and comparing and integrating emerging categories. This process reflects an iterative interplay between deduction and induction—using the theoretically derived dimensions as an initial guide, allowing new categories to emerge from the data, and ultimately integrating both into a coherent ethical framework.
The object-oriented dimension addresses archivers’ responsibilities toward external “others”—what and whom their actions affect. Inductive coding of archival ethics policies yields three object categories that capture layered ethical relations: (1) archived objects, the records and information under care; (2) interactive objects, stakeholders such as creators, users, and other participants; and (3) societal objects, the wider social and institutional contexts in which archiving is embedded. Each category aligns with specific value goals that together define the ethical orientation of this dimension.
If object-oriented ethics specifies the desirable states these “others” should attain, the self-oriented dimension explains how archivers realize them. It focuses on professional virtues and role expectations—who archivers are and how they act—and comprises cultivation & character, knowledge & skills, and strategy & action. Its principles guide conduct and capacity-building rather than prescribing outcomes for others. Together, these subcategories support ethical judgment, integrity, and accountable decision-making across diverse archival situations.
Here, we apply the general ethical framework to web archiving. The object-oriented dimension functions as a diagnostic lens for identifying ethical issues in web archiving practices, while the self-oriented dimension offers guidance for practitioners in strengthening the professional capacities and behavioral norms required to uphold these values in practice. Taken together, the two dimensions provide a comprehensive approach that supports both ethical assessment and capacity building within the field of web archiving.
The object-oriented dimension reveals how ethical tensions in web archiving cluster around three types of objects:
Archived Objects. AI-driven crawling may compromise: Authenticity, due to dynamic content changes or algorithmic reconstruction; Integrity, when page elements fail to load or scripts alter web states; Security, through inadvertent capture of sensitive personal data.
Interactive Objects. Here, ethical tensions relate to: Autonomy, given that users rarely consent to archiving of their online interactions; Privacy, especially where personal data is embedded in social media content; Access, due to uneven representation, capture gaps, or platform restrictions.
Societal Objects. At this level, ethical challenges involve: Justice, as marginalized communities may be under-archived or misrepresented; History & Memory, where archives shape narratives of social events; Diversity & Inclusion, threatened by commercial platform bias; Accountability, as archives can be used for surveillance or misinformation tracking; Sustainability, given the energy demands of large-scale crawling and AI pipelines.
Together, these tensions illustrate the complexity of participation in web archiving: who is included, who is represented, who is implicated, and who has agency.
The self-oriented dimension provides actionable pathways for practitioners, especially in AI-mediated environments.
Compliance. Archivers must navigate diverse legal frameworks, platform terms of service, and algorithmic governance regimes.
Balance. Ethical judgment requires balancing privacy with public interest, autonomy with collective memory, and preservation with harm mitigation.
Transparency. Openness in workflows—including crawler settings, capture frequency, selection criteria, and data processing pipelines—enhances accountability and participatory engagement.
Collaboration. Multi-stakeholder collaboration is essential, involving communities, platform operators, AI developers, researchers, and affected individuals.
Together, these strategies move web archiving from a technically driven activity toward a participatory, socially grounded ethical practice.
Web archiving in the age of AI is a site of negotiation among technologies, institutions, communities, and individuals. This study demonstrates that ethical challenges cannot be reduced to isolated issues but must be organized within a structured, relational, and participatory framework. The proposed two-dimensional structure—object-oriented and self-oriented—serves as both a diagnostic tool and a guide for governance, contributing to more inclusive, responsible, and critically informed digital memory practices. Rather than prescribing fixed rules, it provides a flexible interpretive structure that can be applied to emerging contexts such as AI-driven data extraction, automated curation, platform governance, and large-scale web scraping—all of which complicate participation and responsibility in digital memory work.