{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,22]],"date-time":"2026-10-22T10:40:14Z","timestamp":1792665614111,"version":"4.5.0"},"reference-count":48,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2027,4,1]],"date-time":"2027-04-01T00:00:00Z","timestamp":1806537600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2027,4,1]],"date-time":"2027-04-01T00:00:00Z","timestamp":1806537600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2027,4,1]],"date-time":"2027-04-01T00:00:00Z","timestamp":1806537600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-017"},{"start":{"date-parts":[[2027,4,1]],"date-time":"2027-04-01T00:00:00Z","timestamp":1806537600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2027,4,1]],"date-time":"2027-04-01T00:00:00Z","timestamp":1806537600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-012"},{"start":{"date-parts":[[2027,4,1]],"date-time":"2027-04-01T00:00:00Z","timestamp":1806537600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2027,4,1]],"date-time":"2027-04-01T00:00:00Z","timestamp":1806537600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-004"}],"funder":[{"DOI":"10.13039\/501100001665","name":"Agence nationale de la recherche","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100001665","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100014013","name":"UKRI","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100014013","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Current Opinion in Psychology"],"published-print":{"date-parts":[[2027,4]]},"DOI":"10.1016\/j.copsyc.2026.102478","type":"journal-article","created":{"date-parts":[[2026,10,14]],"date-time":"2026-10-14T15:59:19Z","timestamp":1791993559000},"page":"102478","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":0,"special_numbering":"C","title":["A modular review of the evidence for harmful manipulation by large language models"],"prefix":"10.1016","volume":"74","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4959-188X","authenticated-orcid":false,"given":"Jean-Fran\u00e7ois","family":"Bonnefon","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Carina","family":"Prunkl","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"10.1016\/j.copsyc.2026.102478_bib1","article-title":"International AI safety report 2026","author":"Bengio","year":"2026","journal-title":"arXiv:260221012"},{"key":"10.1016\/j.copsyc.2026.102478_bib2","article-title":"Regulation (EU) 2024\/1689, Article 5: prohibited AI Practices; 2024","year":"2024","journal-title":"Off J Eur Union L"},{"key":"10.1016\/j.copsyc.2026.102478_bib3","doi-asserted-by":"crossref","first-page":"1855","DOI":"10.1038\/s41562-023-01742-2","article-title":"Machine culture","volume":"7","author":"Brinkmann","year":"2023","journal-title":"Nat Hum Behav"},{"key":"10.1016\/j.copsyc.2026.102478_bib4","doi-asserted-by":"crossref","first-page":"653","DOI":"10.1146\/annurev-psych-030123-113559","article-title":"The moral psychology of artificial intelligence","volume":"75","author":"Bonnefon","year":"2024","journal-title":"Annu Rev Psychol"},{"key":"10.1016\/j.copsyc.2026.102478_bib5","article-title":"Lies, damned lies, and distributional language statistics: persuasion and deception with large language models","author":"Jones","year":"2024","journal-title":"arXiv:241217128"},{"key":"10.1016\/j.copsyc.2026.102478_bib6","doi-asserted-by":"crossref","first-page":"354","DOI":"10.1126\/science.adz1697","article-title":"How malicious AI swarms can threaten democracy","volume":"391","author":"Schroeder","year":"2026","journal-title":"Science"},{"key":"10.1016\/j.copsyc.2026.102478_bib7","doi-asserted-by":"crossref","first-page":"126","DOI":"10.1038\/s41586-025-09505-x","article-title":"Delegation to Artificial Intelligence can increase dishonest behaviour","volume":"646","author":"K\u00f6bis","year":"2025","journal-title":"Nature"},{"key":"10.1016\/j.copsyc.2026.102478_bib8","doi-asserted-by":"crossref","first-page":"679","DOI":"10.1038\/s41562-021-01128-2","article-title":"Bad machines corrupt good morals","volume":"5","author":"K\u00f6bis","year":"2021","journal-title":"Nat Hum Behav"},{"key":"10.1016\/j.copsyc.2026.102478_bib9","doi-asserted-by":"crossref","DOI":"10.1126\/science.aec8352","article-title":"Sycophantic AI decreases prosocial intentions and promotes dependence","volume":"391","author":"Cheng","year":"2026","journal-title":"Science"},{"key":"10.1016\/j.copsyc.2026.102478_bib10","article-title":"Simulating psychological risks in Human-AI interactions: Real-Case informed modeling of AI-Induced addiction, Anorexia, depression, homicide, psychosis, and suicide","author":"Archiwaranguprok","year":"2025","journal-title":"arXiv:251108880"},{"key":"10.1016\/j.copsyc.2026.102478_bib11","series-title":"Proceedings of the 2025 ACM conference on fairness, accountability, and transparency","first-page":"599","article-title":"Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providers","author":"Moore","year":"2025"},{"key":"10.1016\/j.copsyc.2026.102478_bib12","doi-asserted-by":"crossref","DOI":"10.1126\/science.adq1814","article-title":"Durably reducing conspiracy beliefs through dialogues with AI","volume":"385","author":"Costello","year":"2024","journal-title":"Science"},{"key":"10.1016\/j.copsyc.2026.102478_bib13","doi-asserted-by":"crossref","first-page":"1164","DOI":"10.1126\/science.ads0433","article-title":"Generative AI as a tool for truth","volume":"385","author":"Bago","year":"2024","journal-title":"Science"},{"key":"10.1016\/j.copsyc.2026.102478_bib14","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3579592","article-title":"Working with AI to persuade: examining a large language model\u2019s ability to generate pro-vaccination messages","volume":"7","author":"Karinshak","year":"2023","journal-title":"Proc ACM Hum-Comput Interact"},{"key":"10.1016\/j.copsyc.2026.102478_bib15","doi-asserted-by":"crossref","first-page":"1176","DOI":"10.1038\/s41558-025-02424-9","article-title":"Using generative AI to increase sceptics' engagement with climate science","volume":"15","author":"Bago","year":"2025","journal-title":"Nat Clim Change"},{"key":"10.1016\/j.copsyc.2026.102478_bib16","series-title":"Manipulatory politics","author":"Goodin","year":"1980"},{"key":"10.1016\/j.copsyc.2026.102478_bib17","doi-asserted-by":"crossref","first-page":"129","DOI":"10.1007\/s10677-006-9052-4","article-title":"What\u2019s wrong with motive manipulation?","volume":"10","author":"Cave","year":"2007","journal-title":"Ethical Theory Moral Pract"},{"key":"10.1016\/j.copsyc.2026.102478_bib18","first-page":"1","article-title":"Online manipulation: hidden influences in a digital world","volume":"4","author":"Susser","year":"2019","journal-title":"Georgetown Law Technology Review"},{"key":"10.1016\/j.copsyc.2026.102478_bib19","series-title":"AI Act regulation (EU) 2024\/1689","year":"2025"},{"key":"10.1016\/j.copsyc.2026.102478_bib20","first-page":"1","article-title":"Technology, autonomy, and manipulation","volume":"8","author":"Susser","year":"2019","journal-title":"Internet Pol Rev"},{"key":"10.1016\/j.copsyc.2026.102478_bib21","series-title":"Manipulation: theory and practice","author":"Coons","year":"2014"},{"key":"10.1016\/j.copsyc.2026.102478_bib22","first-page":"43","article-title":"Manipulative actions: a conceptual and moral analysis","volume":"33","author":"Noggle","year":"1996","journal-title":"Am Philos Q"},{"key":"10.1016\/j.copsyc.2026.102478_bib23","first-page":"37","article-title":"Manipulativeness","volume":"vol. 77","author":"Baron","year":"2003"},{"key":"10.1016\/j.copsyc.2026.102478_bib24","series-title":"Proceedings of the 2018 CHI conference on human factors in computing systems","first-page":"1","article-title":"The dark (patterns) side of UX design","author":"Gray","year":"2018"},{"key":"10.1016\/j.copsyc.2026.102478_bib25","series-title":"Proceedings of the 2021 CHI conference on human factors in computing systems","first-page":"1","article-title":"What makes a dark pattern\u2026dark? Design attributes, normative considerations, and measurement methods","author":"Mathur","year":"2021"},{"key":"10.1016\/j.copsyc.2026.102478_bib26","article-title":"Experimental evidence that conversational artificial intelligence can steer consumer behavior without detection","author":"Werner","year":"2024","journal-title":"arXiv:240912143"},{"key":"10.1016\/j.copsyc.2026.102478_bib27","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2600684123","article-title":"Human preferences are susceptible to covertly misaligned AI advice","volume":"123","author":"Sabour","year":"2026","journal-title":"Proc Natl Acad Sci"},{"key":"10.1016\/j.copsyc.2026.102478_bib28","doi-asserted-by":"crossref","first-page":"281","DOI":"10.1080\/00323187.2024.2335471","article-title":"Large language models can argue in convincing ways about politics, but humans dislike AI authors: implications for governance","volume":"75","author":"Palmer","year":"2023","journal-title":"Polit Sci"},{"key":"10.1016\/j.copsyc.2026.102478_bib29","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2403116121","article-title":"Evaluating the persuasive influence of political microtargeting with large language models","volume":"121","author":"Hackenburg","year":"2024","journal-title":"Proc Natl Acad Sci"},{"key":"10.1016\/j.copsyc.2026.102478_bib30","doi-asserted-by":"crossref","first-page":"6037","DOI":"10.1038\/s41467-025-61345-5","article-title":"LLM-generated messages can persuade humans on policy issues","volume":"16","author":"Bai","year":"2025","journal-title":"Nat Commun"},{"key":"10.1016\/j.copsyc.2026.102478_bib31","first-page":"1","article-title":"On the conversational persuasiveness of GPT-4","author":"Salvi","year":"2025","journal-title":"Nat Hum Behav"},{"key":"10.1016\/j.copsyc.2026.102478_bib32","doi-asserted-by":"crossref","DOI":"10.1126\/science.aea3884","article-title":"The levers of political persuasion with conversational artificial intelligence","volume":"390","author":"Hackenburg","year":"2025","journal-title":"Science"},{"key":"10.1016\/j.copsyc.2026.102478_bib33","first-page":"1","article-title":"Persuading voters using human-artificial intelligence dialogues","author":"Lin","year":"2025","journal-title":"Nature"},{"key":"10.1016\/j.copsyc.2026.102478_bib34","doi-asserted-by":"crossref","DOI":"10.1037\/pspi0000531","article-title":"From people to prompts: understanding susceptibility to persuasion in conversations with artificial intelligence","author":"Tomaino","year":"2026","journal-title":"J Pers Soc Psychol"},{"key":"10.1016\/j.copsyc.2026.102478_bib35","doi-asserted-by":"crossref","DOI":"10.1073\/pnas.2515516122","article-title":"Chatbot Voting Advice Applications inform but seldom sway young unaligned voters","volume":"122","author":"Velez","year":"2025","journal-title":"Proc Natl Acad Sci"},{"key":"10.1016\/j.copsyc.2026.102478_bib36","article-title":"Ads in AI chatbots? An analysis of how large language models navigate conflicts of interest","author":"Wu","year":"2026","journal-title":"arXiv:260408525"},{"key":"10.1016\/j.copsyc.2026.102478_bib37","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3770640","article-title":"Ads that talk back: implications and perceptions of injecting personalized advertising into LLM chatbots","volume":"9","author":"Tang","year":"2025","journal-title":"Proc ACM Interact Mob Wearab Ubiquit Technol"},{"key":"10.1016\/j.copsyc.2026.102478_bib38","series-title":"Proceedings of the 2023 CHI conference on human factors in computing systems","first-page":"1","article-title":"Co-writing with opinionated language models affects users' views","author":"Jakesch","year":"2023"},{"key":"10.1016\/j.copsyc.2026.102478_bib39","article-title":"LLMs as scalable tools for interactive consumer behavior experiments: comparing persuasion strategy effectiveness","author":"Nam","year":"2025","journal-title":"OSF Preprint"},{"key":"10.1016\/j.copsyc.2026.102478_bib40","article-title":"Evaluating frontier models for dangerous capabilities","author":"Phuong","year":"2024","journal-title":"arXiv:240313793"},{"key":"10.1016\/j.copsyc.2026.102478_bib41","article-title":"Evaluating language models for harmful manipulation","author":"Akbulut","year":"2026","journal-title":"arXiv:260325326"},{"key":"10.1016\/j.copsyc.2026.102478_bib42","article-title":"Large language models can effectively convince people to believe conspiracies","author":"Costello","year":"2026","journal-title":"arXiv:260105050"},{"key":"10.1016\/j.copsyc.2026.102478_bib43","unstructured":"This article shows how LLM dialogue can be used to reduce harmful beliefs, providing a useful contrast case for the present review."},{"key":"10.1016\/j.copsyc.2026.102478_bib44","unstructured":"This article provides published evidence that LLM-generated messages can persuade humans on policy issues, while still falling short of harmful manipulation as defined here."},{"key":"10.1016\/j.copsyc.2026.102478_bib45","unstructured":"This article reports the largest behavioral dataset on political persuasion, testing the impact of 19 LLMs on 76,000 participants across 707 issues, and reporting a negligible effect of personalization tactics."},{"key":"10.1016\/j.copsyc.2026.102478_bib46","unstructured":"This article provides published experimental evidence that opinionated language-model assistance can change both what users write and what they later believe, often without users recognizing the system as one-sided."},{"key":"10.1016\/j.copsyc.2026.102478_bib47","unstructured":"A paper from DeepMind that evaluates both the propensity of Gemini 3 to manipulate, and the efficacy of its attempts, across various domains and countries. Gemini does attempt to manipulate beliefs and behaviors, and manages to change beliefs, but has very little impact on behavior."},{"key":"10.1016\/j.copsyc.2026.102478_bib48","unstructured":"This article provides one of the clearest operationalization of a harmful outcome, showing that LLMs can convince participants that a conspiracy theory is true, when they are initially undecided about its truth. It also raises important questions about the ethics of inducing this kind of belief, and how to repair the harm done to participants."}],"container-title":["Current Opinion in Psychology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S2352250X26002174?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S2352250X26002174?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,10,22]],"date-time":"2026-10-22T10:06:05Z","timestamp":1792663565000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S2352250X26002174"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2027,4]]},"references-count":48,"alternative-id":["S2352250X26002174"],"URL":"https:\/\/doi.org\/10.1016\/j.copsyc.2026.102478","relation":{},"ISSN":["2352-250X"],"issn-type":[{"value":"2352-250X","type":"print"}],"subject":[],"published":{"date-parts":[[2027,4]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"A modular review of the evidence for harmful manipulation by large language models","name":"articletitle","label":"Article Title"},{"value":"Current Opinion in Psychology","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.copsyc.2026.102478","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar technologies.","name":"copyright","label":"Copyright"}],"article-number":"102478"}}