EUVIMEDEUVIMED entdecken
Crossref

The levers of political persuasion with conversational artificial intelligence

Kobi Hackenburg iD, Ben M Tappin iD, Luke Hewitt iD, Ed Saunders iD, Sid Black iD, Hause Lin iD, Catherine Fist, Helen Margetts iD, David G Rand iD, Christopher Summerfield iD

Science · 04.12.2025 · 390 (6777)

DOI: 10.1126/science.aea3884 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Science, erschienen am 04.12.2025. Verfasst von Kobi Hackenburg und 9 weiteren Autor:innen. Beteiligte Einrichtung: UK AI Security Institute, London, UK. Hauptthemen laut MEDLINE: Persuasive Communication, Politics, Large Language Models. Die Arbeit wurde bisher 54-mal zitiert (Crossref) und nennt 87 Literaturangaben. Gefördert von: UK Government’s Department for Science, Innovation and Technology.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

There are widespread fears that conversational artificial intelligence (AI) could soon exert unprecedented influence over human beliefs. In this work, in three large-scale experiments ( N = 76,977 participants), we deployed 19 large language models (LLMs)—including some post-trained explicitly for persuasion—to evaluate their persuasiveness on 707 political issues. We then checked the factual accuracy of 466,769 resulting LLM claims. We show that the persuasive power of current and near-future AI is likely to stem more from post-training and prompting methods—which boosted persuasiveness by as much as 51 and 27%, respectively—than from personalization or increasing model scale, which had smaller effects. We further show that these methods increased persuasion by exploiting LLMs’ ability to rapidly access and strategically deploy information and that, notably, where they increased AI persuasiveness, they also systematically decreased factual accuracy.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Persuasive CommunicationPoliticsLarge Language ModelsHumansMale

Autor:innen und Institutionen

  • UK AI Security Institute, London, UK.
  • Oxford Internet Institute, University of Oxford, Oxford, UK.
  • Department of Psychological and Behavioural Science, London School of Economics and Political Science, London, UK.
  • Department of Sociology, Stanford University, Stanford, CA, USA.
  • Sloan School of Management, Massachusetts Institute of Technology, Boston, MA, USA.
  • Department of Information Science, Marketing and Management Communications, and Department of Psychology, Cornell University, Ithaca, NY, USA.

ORCID-Profile: Kobi Hackenburg, Ben M Tappin, Luke Hewitt, Ed Saunders, Sid Black, Hause Lin, Helen Margetts, David G Rand, Christopher Summerfield

Förderung

  • UK Government’s Department for Science, Innovation and Technology

Zitieren

APA

Hackenburg, K., Tappin, B. M., Hewitt, L., Saunders, E., Black, S., Lin, H., Fist, C., Margetts, H., Rand, D. G., & Summerfield, C. (2025). The levers of political persuasion with conversational artificial intelligence. Science, 390(6777). https://doi.org/10.1126/science.aea3884

Vancouver

Hackenburg K, Tappin BM, Hewitt L, Saunders E, Black S, Lin H, et al. The levers of political persuasion with conversational artificial intelligence. Science. 2025;390(6777). doi:10.1126/science.aea3884

BibTeX
@article{Hackenburg2025,
  author = {Hackenburg, Kobi and Tappin, Ben M and Hewitt, Luke and Saunders, Ed and Black, Sid and Lin, Hause and Fist, Catherine and Margetts, Helen and Rand, David G and Summerfield, Christopher},
  title = {The levers of political persuasion with conversational artificial intelligence},
  journal = {Science},
  year = {2025},
  volume = {390},
  number = {6777},
  doi = {10.1126/science.aea3884}
}
Kostenlos starten →