Conversational AI increases political knowledge as effectively as self-directed internet search
Peer-reviewed study from the UK AI Security Institute combining a representative survey of the UK public on the use of conversational AI for political information before the 2024 general election with a series of randomised controlled trials that compare task-directed political conversations with AI against self-directed Google search on belief in true information and misinformation. Chatbot conversations increased political knowledge to the same extent as search across issues, models and prompting strategies.
Publisher
PNAS Nexus (Oxford University Press); UK AI Security Institute
Published
6 Oct 2026
Added
today
Key Findings
- Representative national survey (n = 2,499): in the week before the 2024 UK election, 32% of chatbot users, and 13% of eligible UK voters, had used conversational AI to seek political information relevant to their electoral choice.
- Randomised controlled trials (n = 2,858 in total): task-directed conversations with AI about specific political topics increased political knowledge (higher belief in true information, lower belief in misinformation) to the same extent as self-directed Google search.
- The effect generalised across multiple topics, political perspectives and model families; the authors conclude that the shift toward AI-assisted political information-seeking may not lead to increased public belief in political misinformation.
Methodology Notes
Survey plus RCTs on UK samples; effects are measured on knowledge of true and false claims, not on vote choice or attitudes. Published online 2026-10-06 (PubMed article date; Crossref created 2026-10-06) in PNAS Nexus volume 5 issue 10 (PMID 42840900); the arXiv preprint 2509.05219 was posted 2025-09-08. Verification route: PubMed record with the full abstract and AISI affiliations, the Crossref record and the AISI research page; the Oxford University Press article page blocks crawlers.
Sources
PNAS Nexus article (DOI)(opens in a new tab) (primary)
PubMed record (PMID 42840900)(opens in a new tab) (6 Oct 2026)
arXiv preprint(opens in a new tab) (8 Sept 2025)
UK AI Security Institute research page(opens in a new tab) (8 Sept 2025)
Authors
Lennart Luettgau, Hannah Rose Kirk, Kobi Hackenburg, Jessica Bergs, Henry Davidson, Henry Ogden, Divya Siddarth, Saffron Huang, Christopher Summerfield
Tags
Cite This
APA
Lennart Luettgau et al. (2026). Conversational AI increases political knowledge as effectively as self-directed internet search. PNAS Nexus (Oxford University Press); UK AI Security Institute. https://doi.org/10.1093/pnasnexus/pgag305
Related Insights
People readily follow personal advice from AI but it does not improve their well-being
arXiv (UK AI Security Institute; Limbic AI) · 19 Nov 2025
How is AI impacting people?
UK AI Security Institute (Societal Impacts Team), posted on PsyArXiv · 6 Sept 2026
RealityTest: How People Probe AI Identity and Whether Models Disclose It
arXiv (AI Security Institute, UK Department for Science, Innovation and Technology; University of Oxford) · 29 May 2026
A light-touch AI literacy intervention helps protect against AI political persuasion
arXiv (Carnegie Mellon University, Department of Social and Decision Sciences; Cornell University) · 14 Sept 2026