Google DeepMind researchers

[Skip to main content](https://link.springer.com/article/10.1007/s13347-025-00928-y#main)

## Search

Search by keyword or author

Search

## Navigation

- [Find a journal](https://link.springer.com/journals/)
- [Publish with us](https://www.springernature.com/gp/authors)
- [Track your research](https://link.springernature.com/home/)

# AI’s Epistemic Harm: Reinforcement Learning, Collective Bias, and the New AI Culture War

- Research Article
- [Open access](https://www.springernature.com/gp/open-science/about/the-fundamentals-of-open-access-and-open-research)
- Published: 11 July 2025

- Volume 38, article number 102, (2025)


- [Cite this article](https://link.springer.com/article/10.1007/s13347-025-00928-y#citeas)

You have full access to this [open access](https://www.springernature.com/gp/open-science/about/the-fundamentals-of-open-access-and-open-research) article

[Download PDF](https://link.springer.com/content/pdf/10.1007/s13347-025-00928-y.pdf)

[Save article](https://link.springer.com/article/10.1007/s13347-025-00928-y/save-research?_csrf=m3b7smk1sOD5ssin88E9KZlVedLAk94r)

[View saved research](https://link.springer.com/saved-research)

[![](https://media.springernature.com/w72/springer-static/cover-hires/journal/13347?as=webp)Philosophy & Technology](https://link.springer.com/journal/13347) [Aims and scope](https://link.springer.com/journal/13347/aims-and-scope) [Submit manuscript](https://www.editorialmanager.com/phte)

AI’s Epistemic Harm: Reinforcement Learning, Collective Bias, and the New AI Culture War


[Download PDF](https://link.springer.com/content/pdf/10.1007/s13347-025-00928-y.pdf)

## Abstract

Generative AI is increasingly used as an epistemic tool to aid inquiry, broaden our knowledge and generally help us find things out. It is now embedded in internet search functions to condense information and answer user questions without further need to access websites or news articles. However, models based on generative AI come with inherent flaws which may inhibit their effectiveness as tools to aid responsible inquiry. While the risk of misinformation due to ‘hallucinations’ has been documented, this paper instead focuses on the capacity for these models to present moral or political bias reflective of the preferences of the model designers and crowd worker labellers. This paper takes the novel approach of tracing the ethical issue of bias back to fundamental alignment problem within the process of Reinforced Learning from Human Feedback (RLHF), a key step in training generative AI models. The potential for language models to be influenced by their designers, censored to certain topics, or trained to be more politically leaning one way or another has the capability of insidiously causing epistemic harm. By applying the concepts of _bias of crowds_ to the training process of language models I show how they can impede effective epistemic inquiry and cause epistemic and moral harm. In doing so I present an argument showing that 

... (truncated, 98 KB total)