Call for papers: AI systems for the public interest

Special issue of the Internet Policy Review

assignment_returned Abstract submission deadline: 29 October 2023
assignment_returned Full paper submission deadline: 14 January 2024

 

The number of AI projects claiming to serve the common good or public interest is increasing rapidly (Shi et al., 2021; Cowls, 2021). But often the information on these projects, their initiators, funders, methods, and objectives is not transparent, hindering the goal of serving the public. Against this background arises the necessity for an empirically based discussion on the criteria, processes, and conditions for AI systems to serve the public interest (Floridi et al., 2020; Züger & Asghari, 2022; Bondi et al., 2022). These authors implicitly or explicitly build on working definitions of public interest, such as the one provided by Bozeman (2007), who defines it as “the outcomes best serving the long-run survival and well-being of a social collective construed as a ‘public’”. The public interest, according to Bozeman, must be negotiated case by case by the ones affected—and cannot be defined universally or for all times. The public interest, according to most accounts, is incompatible with pure profit-maximisation goals.

Building on this understanding of the public interest, a number of implications arise for the development and implementation of AI systems. These are formulated slightly differently by various scholars (e.g., Züger et al., 2022; Blankertz & Kelch, 2023), but often include the following aspects:

  • a justification for the use of the system in the public interest

  • transparency and public deliberation

  • participation or participatory design

  • avoiding negative externalities / serving equality and equity

  • access and openness of the system and of data for evaluation, validation and re-usability

  • technical standards and safeguards

  • sustainability of the system and the organisation that hosts it

Many AI projects touch upon sensitive areas with public wellbeing at stake, such as public health, mobility or the justice system, while most often leave aside how they meet these aspects mentioned above to safeguard the interests of the public. Also, the impact of public interest AI projects is hardly ever evaluated and discussed, in particular in a holistic and transparent manner.

Given the lack of detailed data on AI for public interest projects, their methods and principles of development do not have shared standards. From a research but also a policy perspective, this lack of empirical data to analyse what kind of criteria and which actors define such projects in practice is problematic. Discussing the idea of public interest AI without any criteria and theoretical background hollows it out and may even turn it into a whitewashing label.

Scope of special issue

In this interdisciplinary special issue we aim to connect public interest theory (Bozeman, 2007; Feintuck, 2002; Held, 1970) to the debate about AI projects and foster exchange amongst existing projects that use AI to serve the public interest to explore common challenges, methods, and standards. Our goal is to identify, through case studies and theoretical works, a common ground of shared standards for the democratic procedure of development and evaluation of public interest AI.

We hope this special issue can contribute to an exchange of recent empirical and conceptual research findings on AI systems serving the public interest and the economic, organisational and technological conditions that underpin their impact.

We are particularly interested in factors (which include standards and legal instruments) that are relevant to support the flourishing of public interest AI, as well as to preventing outcomes that hinder or hurt the public interest.

We encourage submissions that report on and critically examine work in progress, case studies, tools, or present a synthesis of empirical insights on AI in the public interest. We will be especially attentive to case studies from a European context or other world regions that relate and enrich the debate. Theoretical works are also welcome, as well as submissions on practices in the public interest AI context, including issues around data collection, data sharing, auditing and other aspects of the AI lifecycle.

Topics of interest include:

  • AI for accessibility

  • AI for education

  • AI for equality & equitable AI (including intersectional and feminist perspectives)

  • AI for public debate & journalism

  • AI for public health & medicine

  • AI for public infrastructure & mobility

  • AI for public safety

  • AI for public services

  • AI for sustainability

We also specifically welcome works that focus on the tensions that arise for public interest AI, such as:

  • the tension between commercial goals and the public interest

  • the tension between a research agenda and public interest goals

  • the tension arising from conflicting interests, which can respectively be argued to stand for a public interest

  • the tension between Euro-American developers and developers and interest in other regions of the world

All papers need to specifically connect to public interest theory. Submissions can contribute to the theory in general, or investigate particular cases of AI for the public interest. They can expand or contradict the existing knowledge on public interest AI theory, but the connection to public interest theory must be clearly discussed. Theoretical submissions need to engage with AI practices, while conferring about the relation of public interest theory to other conceptually relevant issues.

Special issue editors

  • Theresa Züger (Alexander von Humboldt Institute for Internet and Society) zueger@hiig.de (primary contact person)

  • Hadi Asghari (Alexander von Humboldt Institute for Internet and Society) hadi.asghari@hiig.de

Important dates

  • Submission of 750-1000 word abstracts by October 29, 2023 (AOE). Submissions should be emailed to piai@hiig.de (please put “Special issue” in the subject line). The abstract should articulate:

    • the issue or research question to be discussed,

    • the case study on which the article builds,

    • the methodological or critical framework used,

    • an indication of the expected findings or conclusions,

    • five key references.

  • Notification of authors regarding abstracts: 24 November 2023

  • Full paper submissions deadline: 14 January 2024

  • Comprehensive peer-reviewer aimed for 1 April 2024

  • The planned publication date of this special issue is Q3 2024.

If you have any questions, please don’t hesitate to contact the special issue editors via piai@hiig.de. For more information, visit our website at https://www.hiig.de/en/project/public-interest-ai/ or https://publicinterst.ai

References

Blankertz, A., Kelch, F. (2023). “Eight requirements: Making digital policy serve the public interest”. Wikimedia Deutschland. https://upload.wikimedia.org/wikipedia/commons/a/a8/Brochure_Eight_requirements._Making_digital_policy_serve_the_public_interest.pdf

Bondi, E, Xu, L, Acosta-Navas, D., & Killian, J.A. (2021). “Envisioning Communities: A Participatory Approach Towards AI for Social Good”. Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society. https://doi.org/10.1145/3461702.3462612.

Bozeman, B. (2007). “Public Values and Public Interest: Counterbalancing Economic Individualism”. Georgetown University Press: Washington, D.C.

Cowls, J. (2021). “‘AI for Social Good’: Whose Good and Who’s Good?”. Philosophy & Technology 34. https://doi.org/10.1007/s13347-021-00466-3

Feintuck, M. (2004). “‘The Public Interest’ in Regulation”. Oxford University Press: Oxford.

Floridi, L, Cowls, J., King, T.C., Taddeo, M. (2020). “How to Design AI for Social Good: Seven Essential Factors”. Science and Engineering Ethics, 26:1771-1796. https://doi.org/10.1007/s11948-020-00213-5

Held, V. (1970). “The Public Interest and Individual Interests”. Basic Books: New York.

Shi, Z. R., Wang, C., & Fang, F. (2020). “Artificial Intelligence for Social Good: A Survey”. arXiv Preprint [cs]. http://arxiv.org/abs/2001.01818.

Züger, T., Asghari, H. (2022). “AI for the public. How public interest theory shifts the discourse on AI”. AI & Soc 38, 815–828. https://doi.org/10.1007/s00146-022-01480-5

Züger, T., Faßbender, J., Kuper, F., Nenno, S., Katzy-Reinshagen, A., & Kühnlein, I. (2022). “Civic Coding: Grundlagen und empirische Einblicke zur Unterstützung gemeinwohlorientierter KI”, hrsg im Rahmen der Initiative Civic Coding vom Bundesministerium für Umwelt, Naturschutz, nukleare Sicherheit und Verbraucherschutz, Bundesministerium für Arbeit und Soziales, Bundesministerium für Familie, Senioren, Frauen und Jugend, Berlin https://www.civic-coding.de/fileadmin/civic-ai/Dateien/Civic_Coding_Forschungsbericht.pdf