Mussgnug, Alexander and Leonelli, Sabina and Vallor, Shannon (2026) Responsible AI for Research: Between Moral Philosophy and Philosophy of Science.

Text Responsible AI in Research.pdf Download (249kB)

Abstract

Today, scientists employ AI to identify relevant literature, facilitate large-scale data analysis, and automate routine experimental procedures. Tomorrow, fully autonomous AI may fundamentally alter the way science is done (Royal Society 2024). Current and future applications of AI in science have motivated a wide range of research exploring how to responsibly integrate AI into scientific practice.
While AI undoubtedly opens up enticing technical possibilities, worries about a decline in quality of scientific procedures and outputs, combined with the increasing inscrutability of research systems, have raised concerns regarding the scientific implications of using AI. The social and ethical implications of such developments are also cause for concern, not least because scientific research often grounds far-reaching political, professional, and personal decision-making. The automation of key moments of evaluation and judgement within the scientific process thus leads to critical questions regarding quality, trust, and accountability beyond the realm of scientific research alone.
The scientific and social dimensions of AI’s impact on science are, in our view, inextricably linked. Any problems in the quality and reliability of methods and outputs will reflect on misleading, problematic or downright damaging uses of science in society. Conversely, the ongoing erosion of trust and accountability through AI and the growing resistance toward socially harmful uses of the technology, will ultimately also shape the very conditions for how AI is implemented into scientific research. While philosophers can play a central role in analysing these interrelated issues and their repercussions, discussions around the ethical and epistemic dimensions of AI in science tend to be siloed into different domains of philosophy.
On the one hand, the epistemic dimensions of AI applications are most comprehensively discussed in the philosophy of science. Among others, philosophers of science have explored how the opacity of deep learning models impacts scientific discovery (e.g., Boge 2022), outlined under what conditions AI models lead to robust results (e.g., Freiesleben & Grote 2023), and debated the potential of fully automated knowledge generation through AI (e.g., Bertolaso & Sterpetti 2020). On the other hand, ethical considerations surrounding the development and use of AI are most extensively explored in the moral philosophy of technology. Within it, scholars have investigated issues such as the fairness and ecological sustainability of AI applications (e.g., Selbst et al 2019; van Wynsberghe 2021), the degree of moral responsibility attributable to AI developers (e.g., Oimann & Tollon 2025), as well as the general effects of AI automation on our social, political, and moral lives (e.g., Vallor 2024).
Our chapter contributes to ongoing research attempting to overcome this division by linking debates in the history and philosophy of science with considerations from the ethics of technology. Section two introduces existing scholarship that connects epistemic and ethical implications of AI. Section three presents our case study. We comment on the background conditions of AI in science, and particularly on questions of inclusivity and equity. In so doing, we provide brief examples of the close connection of epistemic and ethical concerns around the use of AI for research, drawing from scholarship in both the philosophy of science and the moral philosophy of technology. We conclude in section four.

---

Export/Citation:

EndNote | BibTeX | Dublin Core | ASCII/Text Citation (Chicago) | HTML Citation | OpenURL

Social Networking:

Share |

---

Item Type:	Published Article or Volume
Creators:	Creators Email ORCID Mussgnug, Alexander alexander@mussgnug.de 0000-0002-5951-057X Leonelli, Sabina sabina.leonelli@tum.de 0000-0002-7815-6609 Vallor, Shannon svalllor@ed.ac.uk 0000-0001-7036-5222
Keywords:	AI ethics, ethics of science, responsible AI
Subjects:	General Issues > Data Specific Sciences > Artificial Intelligence General Issues > Ethical Issues General Issues > Technology
Depositing User:	Mr Alexander Mussgnug
Date Deposited:	15 May 2026 19:30
Last Modified:	15 May 2026 19:30
Item ID:	29639
Subjects:	General Issues > Data Specific Sciences > Artificial Intelligence General Issues > Ethical Issues General Issues > Technology
Date:	2026
URI:	https://philsci-archive.pitt.edu/id/eprint/29639

Monthly Views for the past 3 years

Monthly Downloads for the past 3 years

Plum Analytics

Actions (login required)

View Item