{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,16]],"date-time":"2026-10-16T17:56:51Z","timestamp":1792173411918,"version":"4.4.0"},"reference-count":53,"publisher":"National Academy of Sciences","issue":"42","license":[{"start":{"date-parts":[[2026,10,16]],"date-time":"2026-10-16T00:00:00Z","timestamp":1792108800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":["www.pnas.org"],"crossmark-restriction":true},"short-container-title":["Proc. Natl. Acad. Sci. U.S.A."],"published-print":{"date-parts":[[2026,10,20]]},"abstract":"<jats:p>\n                    There is substantial concern about the ability of advanced AI to influence people\u2019s behavior. A rapidly growing body of research has found that AI can produce large persuasive effects on people\u2019s attitudes, but whether AI can persuade people to take consequential real-world actions has remained unclear. In two large preregistered experiments (\n                    <jats:inline-formula>\n                      <mml:math xmlns:mml=\"http:\/\/www.w3.org\/1998\/Math\/MathML\" display=\"inline\" overflow=\"scroll\">\n                        <mml:mrow>\n                          <mml:mi>N<\/mml:mi>\n                          <mml:mo>=<\/mml:mo>\n                          <mml:mn>17<\/mml:mn>\n                          <mml:mo>,<\/mml:mo>\n                          <mml:mn>950<\/mml:mn>\n                        <\/mml:mrow>\n                      <\/mml:math>\n                    <\/jats:inline-formula>\n                    responses from 14,779 people), we used conversational AI models to persuade participants on a range of attitudinal and behavioral outcomes, including signing real petitions and donating money to charity. We found sizable AI persuasion effects on these behavioral outcomes (e.g.\n                    <jats:inline-formula>\n                      <mml:math xmlns:mml=\"http:\/\/www.w3.org\/1998\/Math\/MathML\" display=\"inline\" overflow=\"scroll\">\n                        <mml:mo>+<\/mml:mo>\n                      <\/mml:math>\n                    <\/jats:inline-formula>\n                    19.7 percentage points on petition signing). However, we observed no evidence of a correlation between AI persuasion effects on attitudes and effects on behavior. Moreover, we replicated prior findings that information provision drove effects on attitudes, but found no such evidence for our behavioral outcomes. In a test of eight behavioral persuasion strategies, the strategy focused on information provision produced the smallest behavioral effect, though differences among the eight were small. Taken together, these results suggest that previous findings relying on attitudinal outcomes may generalize poorly to behavior, and therefore risk substantially mischaracterizing the real-world behavioral impact of AI persuasion.\n                  <\/jats:p>","DOI":"10.1073\/pnas.2611368123","type":"journal-article","created":{"date-parts":[[2026,10,16]],"date-time":"2026-10-16T17:39:04Z","timestamp":1792172344000},"update-policy":"https:\/\/doi.org\/10.1073\/pnas.cm10313","source":"Crossref","is-referenced-by-count":0,"title":["AI can persuade people to take political actions"],"prefix":"10.1073","volume":"123","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-2071-1726","authenticated-orcid":false,"given":"Kobi","family":"Hackenburg","sequence":"first","affiliation":[{"id":[{"id":"https:\/\/ror.org\/052gg0110","id-type":"ROR","asserted-by":"publisher"}],"name":"Oxford Internet Institute, University of Oxford","place":["Oxford, United Kingdom"]},{"name":"UK AI Security Institute","place":["London, United Kingdom"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Luke","family":"Hewitt","sequence":"additional","affiliation":[{"name":"UK AI Security Institute","place":["London, United Kingdom"]},{"id":[{"id":"https:\/\/ror.org\/00f54p054","id-type":"ROR","asserted-by":"publisher"}],"name":"Department of Sociology, Stanford University","place":["Stanford"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-6675-9294","authenticated-orcid":false,"given":"Caroline","family":"Wagner","sequence":"additional","affiliation":[{"name":"UK AI Security Institute","place":["London, United Kingdom"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2941-2653","authenticated-orcid":false,"given":"Christopher","family":"Summerfield","sequence":"additional","affiliation":[{"id":[{"id":"https:\/\/ror.org\/052gg0110","id-type":"ROR","asserted-by":"publisher"}],"name":"Oxford Internet Institute, University of Oxford","place":["Oxford, United Kingdom"]},{"name":"UK AI Security Institute","place":["London, United Kingdom"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3826-8016","authenticated-orcid":false,"given":"Ben M.","family":"Tappin","sequence":"additional","affiliation":[{"id":[{"id":"https:\/\/ror.org\/0090zs177","id-type":"ROR","asserted-by":"publisher"}],"name":"The Department of Psychological and Behavioural Science, London School of Economics and Political Science","place":["London, United Kingdom"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"341","published-online":{"date-parts":[[2026,10,16]]},"reference":[{"key":"e_1_3_4_1_2","doi-asserted-by":"publisher","DOI":"10.1007\/s13347-024-00756-6"},{"key":"e_1_3_4_2_2","unstructured":"M. Burtell T. Woodside Artificial influence: An analysis of ai-driven persuasion. arXiv [Preprint] (2023). https:\/\/arxiv.org\/abs\/2303.08721 (Accessed 12 March 2026)."},{"key":"e_1_3_4_3_2","unstructured":"Z. Chen et al. A framework to assess the persuasion risks large language model chatbots pose to democratic societies. arXiv [Preprint] (2025). https:\/\/arxiv.org\/abs\/2505.00036 (Accessed 12 March 2026)."},{"key":"e_1_3_4_4_2","unstructured":"S. El-Sayed et al. A mechanism-based approach to mitigating harms from persuasive generative AI. arXiv [Preprint] (2024). https:\/\/arxiv.org\/abs\/2404.15058 (Accessed 12 March 2026)."},{"key":"e_1_3_4_5_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41562-025-02309-z"},{"key":"e_1_3_4_6_2","doi-asserted-by":"publisher","DOI":"10.1126\/science.aea3884"},{"key":"e_1_3_4_7_2","doi-asserted-by":"publisher","DOI":"10.1093\/pnasnexus\/pgae034"},{"key":"e_1_3_4_8_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-61345-5"},{"key":"e_1_3_4_9_2","doi-asserted-by":"publisher","DOI":"10.1126\/science.adq1814"},{"key":"e_1_3_4_10_2","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2413443122"},{"key":"e_1_3_4_11_2","unstructured":"P. Schoenegger et al. Large language models are more persuasive than incentivized human persuaders. arXiv [Preprint] (2025). https:\/\/arxiv.org\/abs\/2505.09662 (Accessed 12 March 2026)."},{"key":"e_1_3_4_12_2","doi-asserted-by":"publisher","DOI":"10.1093\/pnasnexus\/pgae035"},{"key":"e_1_3_4_13_2","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2412815122"},{"key":"e_1_3_4_14_2","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1416587112"},{"key":"e_1_3_4_15_2","doi-asserted-by":"publisher","DOI":"10.1093\/joc\/jqab017"},{"key":"e_1_3_4_16_2","doi-asserted-by":"publisher","DOI":"10.1017\/psrm.2014.10"},{"key":"e_1_3_4_17_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41562-023-01813-4"},{"key":"e_1_3_4_18_2","doi-asserted-by":"publisher","DOI":"10.1093\/restud\/rdx033"},{"key":"e_1_3_4_19_2","doi-asserted-by":"publisher","DOI":"10.1037\/a0020761"},{"key":"e_1_3_4_20_2","doi-asserted-by":"publisher","DOI":"10.2307\/2585837"},{"key":"e_1_3_4_21_2","doi-asserted-by":"publisher","DOI":"10.1177\/1532673X09331613"},{"key":"e_1_3_4_22_2","doi-asserted-by":"publisher","DOI":"10.1080\/17457289.2016.1270288"},{"key":"e_1_3_4_23_2","unstructured":"C. Akbulut Evaluating language models for harmful manipulation. arXiv [Preprint] (2026). https:\/\/doi.org\/10.48550\/arXiv.2603.25326 (Accessed 12 March 2026)."},{"key":"e_1_3_4_24_2","doi-asserted-by":"publisher","DOI":"10.1086\/698714"},{"key":"e_1_3_4_25_2","doi-asserted-by":"publisher","DOI":"10.1007\/s11109-015-9308-6"},{"key":"e_1_3_4_26_2","doi-asserted-by":"crossref","unstructured":"B. M. Tappin For digital mass persuasion exposure matters more than persuasiveness. psyarxiv [Preprint] (2025). https:\/\/doi.org\/10.31234\/osf.io\/5ewbu_v1 (Accessed 12 March 2026).","DOI":"10.31234\/osf.io\/5ewbu_v1"},{"key":"e_1_3_4_27_2","doi-asserted-by":"publisher","DOI":"10.1037\/hea0000294"},{"key":"e_1_3_4_28_2","doi-asserted-by":"publisher","DOI":"10.1017\/S000305540808009X"},{"key":"e_1_3_4_29_2","doi-asserted-by":"publisher","DOI":"10.1037\/0033-2909.132.2.249"},{"key":"e_1_3_4_30_2","unstructured":"K. Hackenburg AI systems out-persuade expert humans. arXiv [Preprint] (2026). https:\/\/doi.org\/10.48550\/arXiv.2606.16475 (Accessed 3 July 2026)."},{"key":"e_1_3_4_31_2","doi-asserted-by":"publisher","DOI":"10.2307\/1914185"},{"key":"e_1_3_4_32_2","doi-asserted-by":"publisher","DOI":"10.3758\/s13428-021-01694-3"},{"key":"e_1_3_4_33_2","doi-asserted-by":"crossref","unstructured":"M. N. Stagnaro Representativeness and response validity across nine opt-in online samples. Nat. Hum. Behav. 10 1441\u20131454 (2026).","DOI":"10.1038\/s41562-026-02438-z"},{"key":"e_1_3_4_34_2","doi-asserted-by":"publisher","DOI":"10.1093\/joc\/jqae042"},{"key":"e_1_3_4_35_2","volume-title":"Altruism in Humans","author":"Batson C. D.","year":"2011","unstructured":"C. D. Batson, Altruism in Humans (Oxford University Press, New York, ed. 1, 2011).","edition":"1"},{"key":"e_1_3_4_36_2","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9221.2007.00563.x"},{"key":"e_1_3_4_37_2","doi-asserted-by":"publisher","DOI":"10.1348\/014466601164939"},{"key":"e_1_3_4_38_2","doi-asserted-by":"publisher","DOI":"10.1080\/10463283.2024.2334563"},{"key":"e_1_3_4_39_2","doi-asserted-by":"publisher","DOI":"10.1111\/spc3.12265"},{"key":"e_1_3_4_40_2","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.1103343108"},{"key":"e_1_3_4_41_2","doi-asserted-by":"publisher","DOI":"10.1111\/cdev.12244"},{"key":"e_1_3_4_42_2","doi-asserted-by":"publisher","DOI":"10.1037\/h0023552"},{"key":"e_1_3_4_43_2","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0197066"},{"key":"e_1_3_4_44_2","doi-asserted-by":"publisher","DOI":"10.1146\/annurev.psych.55.090902.142015"},{"key":"e_1_3_4_45_2","doi-asserted-by":"publisher","DOI":"10.1348\/014466603322595248"},{"key":"e_1_3_4_46_2","doi-asserted-by":"publisher","DOI":"10.1348\/014466607X258704"},{"key":"e_1_3_4_47_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-1-4612-4964-1"},{"key":"e_1_3_4_48_2","doi-asserted-by":"publisher","DOI":"10.7208\/chicago\/9780226821832.001.0001"},{"key":"e_1_3_4_49_2","doi-asserted-by":"publisher","DOI":"10.1037\/0033-2909.134.4.504"},{"key":"e_1_3_4_50_2","doi-asserted-by":"publisher","DOI":"10.1037\/xge0000563"},{"key":"e_1_3_4_51_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.obhdp.2012.08.002"},{"key":"e_1_3_4_52_2","doi-asserted-by":"publisher","DOI":"10.1093\/poq\/nfr025"},{"key":"e_1_3_4_53_2","unstructured":"K. Hackenburg L. Hewitt C. Wagner C. Summerfield B. M. Tappin Data and code for \u201cAI can persuade people to take political actions.\u201d GitHub. https:\/\/github.com\/kobihackenburg\/AI-action-persuasion. Deposited 30 March 2026."}],"container-title":["Proceedings of the National Academy of Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pnas.org\/doi\/pdf\/10.1073\/pnas.2611368123","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,16]],"date-time":"2026-10-16T17:39:22Z","timestamp":1792172362000},"score":1,"resource":{"primary":{"URL":"https:\/\/pnas.org\/doi\/10.1073\/pnas.2611368123"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,16]]},"references-count":53,"journal-issue":{"issue":"42","published-print":{"date-parts":[[2026,10,20]]}},"alternative-id":["10.1073\/pnas.2611368123"],"URL":"https:\/\/doi.org\/10.1073\/pnas.2611368123","relation":{},"ISSN":["0027-8424","1091-6490"],"issn-type":[{"value":"0027-8424","type":"print"},{"value":"1091-6490","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,16]]},"assertion":[{"value":"2026-04-03","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-09-15","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-10-16","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e2611368123"}}