-
The making of digital ghosts: designing ethical AI afterlives
The rapid proliferation of AI-mediated digital afterlife technologies, from chatbots trained on personal data to voice clones and posthumous avatars, has generated a substantial body of ethical literature identifying the moral risks of posthumous simulation. Yet this growing consensus has not been matched by frameworks capable of translating ethical principles into operational design constraints. This…
-
Misinformation, Disinformation, and Autonomy. Against the Epistemic Paternalism Trap @ Brocher Foundation 2026
Most tools developed to counter the issues of misinformation and disinformation at scale replicate the very problem they aim to solve. By filtering, flagging, and moderating content, such tools replace human judgment with algorithmic judgment rather than empowering the people navigating the information ecosystem. This is what we refer to as the epistemic paternalism trap.…
-
Source framing triggers systematic bias in large language models
Abstract Large language models (LLMs) are increasingly used to evaluate text, raising urgent questions about whether their judgments are consistent, unbiased, and robust to framing effects. Here, we examine inter- and intramodel agreement across four state-of-the-art LLMs tasked with evaluating 4800 narrative statements on 24 different topics of social, political, and public health relevance, for…
-
S.A.R.A.H. and the decline of trust in health information: a case study
This case study critically examines S.A.R.A.H. (Smart AI Resource Assistant for Health) as an element of the World Health Organization’s (WHO) digital health strategy, focusing on its design and generated content. Launched in April 2024 to provide accessible health information and combat misinformation, S.A.R.A.H. utilizes generative AI to engage users across diverse health topics. Despite…
-
Emotional prompting amplifies disinformation generation in AI large language models
Introduction: The emergence of artificial intelligence (AI) large language models (LLMs), which can produce text that closely resembles human-written content, presents both opportunities and risks. While these developments offer significant opportunities for improving communication, such as in health-related crisis communication, they also pose substantial risks by facilitating the creation of convincing fake news and disinformation.…
-
The role of AI in election interference and political propaganda in Europe
Notes for the panel discussion “Artificial Intelligence: A double-edged sword for truth” @ Voices 2025 Introduction Existing and emerging digital technologies pose significant concerns, challenges, and risks for the functioning of healthy and full democracies (see: EGE 2023). AI-driven disinformation and misinformation pose a growing threat to society’s information ecosystem. Advances in generative AI enable…
-
Evaluating Transformers on the Ethical Question of Euthanasia
We evaluate the ability of transformers to test if they can reliably predict the recom–mendation by a euthanasia commission on whether, given the report of the commission, euthanasia was justified. As evaluation strategy we prompt the transformer to give binary yes/no answer in chatbot mode, which we then compare to the final verdict on the…
-
The Dual Nature of AI in Information Dissemination: Ethical Considerations
Infodemics pose significant dangers to public health and to the societal fabric, as the spread of misinformation can have far-reaching consequences. While artificial intelligence (AI) systems have the potential to craft compelling and valuable information campaigns with positive repercussions for public health and democracy, concerns have arisen regarding the potential use of AI systems to…
-
Disruptive Technologies and Open Science: How Open Should Open Science Be? A ‘Third Bioethics’ Ethical Framework
This paper investigates the ethical implications of applying open science (OS) practices on disruptive technologies, such as generative AIs. Disruptive technologies, characterized by their scalability and paradigm-shifting nature, have the potential to generate significant global impact, and carry a risk of dual use. The tension arises between the moral duty of OS to promote societal…
-
Large language models at the crossroads of life, death and data in assisted suicide cases @ Digital Bioethics. Hannover 2024
I presented these slides at the Digital Bioethics Symposium organized by Frank Ursin and Sabine Salloch – Medizinische Hochschule Hannover (MHH) and funded by Volkswagen Stiftung. See also: https://renkulab.io/projects/giovanni.spitale/as-case-generation (code presented in the slides)
