The ghost in the machine speaks with an American accent: cultural value drift in early GPT-3 and the case for pluralist evaluation of generative AI

Journal article

Early large language models (LLMs) were released with minimal alignment, offering a rare view of how generative systems reframe the ethical values embedded in human texts. We examine outputs from a 2021 version of OpenAI’s base GPT-3, prompting it to summarise culturally diverse source materials (laws, political speeches, and philosophical works) and interpreting results through a descriptive, moral value pluralist lens. Where possible, we contextualise outputs with cross-national datasets such as the World Values Survey. We document recurring value drift: Australia’s firearm policy is recast as a threat to liberty; de Beauvoir’s feminist critique becomes gender-essentialist dating advice; and Merkel’s humanitarian appeal is recast as immigration control. In contrast, multilateral documents (UN/UNESCO) exhibit greater value stability, suggesting consensus-crafted language can buffer against cultural mutation. We argue that these early behaviours (observed before extensive fine-tuning and safety layers) provide a baseline for understanding how training distributions shape normative framing. Our contribution is twofold: (1) empirical evidence that value drift can invert or overwrite encoded values along predictable cultural axes, and (2) a pluralist, descriptive evaluation method that surfaces whose values dominate and when. We conclude with implications for culturally inclusive evaluation and alignment in contemporary LLMs.

Authors: Johnson, R; Dias Duran, LD; Panai, E; Menéndez González, N; Pistilli, G

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Springer
• Journal: AI and Ethics
• Volume: 6
• Issue: 2
• Article number: 212
• Publication date: 2026-03-23
• Acceptance date: 2026-02-05
• DOI: 10.1007/s43681-026-01038-x
• EISSN: 2730-5961
• ISSN: 2730-5953
• Language: English
• Keywords: World values survey; Evaluating AI; Cultural bias; Generative AI; Moral value pluralism; Aligning AI