No Cover Image

Journal article 213 views 25 downloads

AI-generated images of familiar faces are indistinguishable from real photographs

Robin S. S. Kramer Orcid Logo, Alex Jones Orcid Logo, Daniel Fitousi, Jeremy Tree Orcid Logo

Cognitive Research: Principles and Implications, Volume: 10, Issue: 1

Swansea University Authors: Alex Jones Orcid Logo, Jeremy Tree Orcid Logo

  • 70692.VoR.pdf

    PDF | Version of Record

    © The Author(s) 2025. This article is licensed under a Creative Commons Attribution 4.0 International License.

    Download (1.82MB)

Check full text

DOI (Published version): 10.1186/s41235-025-00683-w

Abstract

Human users are now able to generate synthetic face images with artificial intelligence (AI) tools. Although indistinguishable from real photographs, these images have tended to feature fictional identities that do not exist in the real world. As a result, their use in applied contexts, including th...

Full description

Published in: Cognitive Research: Principles and Implications
ISSN: 2365-7464
Published: Springer Science and Business Media LLC 2025
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa70692
first_indexed 2025-10-15T22:02:10Z
last_indexed 2025-11-15T14:33:37Z
id cronfa70692
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2025-11-14T12:50:22.9035411</datestamp><bib-version>v2</bib-version><id>70692</id><entry>2025-10-15</entry><title>AI-generated images of familiar faces are indistinguishable from real photographs</title><swanseaauthors><author><sid>a24e1e2a89b0a9120fe03b481a629edd</sid><ORCID>0000-0003-3600-3644</ORCID><firstname>Alex</firstname><surname>Jones</surname><name>Alex Jones</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>373fd575114a743d502a979c6161b1ad</sid><ORCID>0000-0001-6000-8125</ORCID><firstname>Jeremy</firstname><surname>Tree</surname><name>Jeremy Tree</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2025-10-15</date><deptcode>PSYS</deptcode><abstract>Human users are now able to generate synthetic face images with artificial intelligence (AI) tools. Although indistinguishable from real photographs, these images have tended to feature fictional identities that do not exist in the real world. As a result, their use in applied contexts, including the spread of fake information, is similarly limited. Here, we investigated a new method for generating face images (via ChatGPT plus DALL&#xB7;E) and its application to both fictional and real (in this case, celebrity) identities. Our results demonstrated that generated images of both fictional (Experiment 1) and celebrity identities (Experiment 2) could not be distinguished from real photographs. Further, providing additional real photographs for comparison during the task resulted in limited gains (Experiments 3 and 4). Finally, prior familiarity with celebrity faces produced only modest performance improvements. Therefore, new methods of detection should be explored as a matter of urgency since the latest &#x2018;off the shelf&#x2019; AI tools can now generate face images of real people that are essentially undetectable as synthetic to most human observers.</abstract><type>Journal Article</type><journal>Cognitive Research: Principles and Implications</journal><volume>10</volume><journalNumber>1</journalNumber><paginationStart/><paginationEnd/><publisher>Springer Science and Business Media LLC</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic>2365-7464</issnElectronic><keywords>Face perception; Deepfakes; ChatGPT; Artificial intelligence</keywords><publishedDay>14</publishedDay><publishedMonth>10</publishedMonth><publishedYear>2025</publishedYear><publishedDate>2025-10-14</publishedDate><doi>10.1186/s41235-025-00683-w</doi><url/><notes/><college>COLLEGE NANME</college><department>Psychology School</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>PSYS</DepartmentCode><institution>Swansea University</institution><apcterm>Another institution paid the OA fee</apcterm><funders>This work was supported by the Israel Science Foundation grant (ISF-1498/21) to DF.</funders><projectreference/><lastEdited>2025-11-14T12:50:22.9035411</lastEdited><Created>2025-10-15T20:20:29.5972451</Created><path><level id="1">Faculty of Medicine, Health and Life Sciences</level><level id="2">School of Psychology</level></path><authors><author><firstname>Robin S. S.</firstname><surname>Kramer</surname><orcid>0000-0001-8339-8832</orcid><order>1</order></author><author><firstname>Alex</firstname><surname>Jones</surname><orcid>0000-0003-3600-3644</orcid><order>2</order></author><author><firstname>Daniel</firstname><surname>Fitousi</surname><order>3</order></author><author><firstname>Jeremy</firstname><surname>Tree</surname><orcid>0000-0001-6000-8125</orcid><order>4</order></author></authors><documents><document><filename>70692__35634__eb5820ee81f848e296a3a357e1d11381.pdf</filename><originalFilename>70692.VoR.pdf</originalFilename><uploaded>2025-11-14T12:47:42.6710431</uploaded><type>Output</type><contentLength>1910872</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>&#xA9; The Author(s) 2025. This article is licensed under a Creative Commons Attribution 4.0 International License.</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>http://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling 2025-11-14T12:50:22.9035411 v2 70692 2025-10-15 AI-generated images of familiar faces are indistinguishable from real photographs a24e1e2a89b0a9120fe03b481a629edd 0000-0003-3600-3644 Alex Jones Alex Jones true false 373fd575114a743d502a979c6161b1ad 0000-0001-6000-8125 Jeremy Tree Jeremy Tree true false 2025-10-15 PSYS Human users are now able to generate synthetic face images with artificial intelligence (AI) tools. Although indistinguishable from real photographs, these images have tended to feature fictional identities that do not exist in the real world. As a result, their use in applied contexts, including the spread of fake information, is similarly limited. Here, we investigated a new method for generating face images (via ChatGPT plus DALL·E) and its application to both fictional and real (in this case, celebrity) identities. Our results demonstrated that generated images of both fictional (Experiment 1) and celebrity identities (Experiment 2) could not be distinguished from real photographs. Further, providing additional real photographs for comparison during the task resulted in limited gains (Experiments 3 and 4). Finally, prior familiarity with celebrity faces produced only modest performance improvements. Therefore, new methods of detection should be explored as a matter of urgency since the latest ‘off the shelf’ AI tools can now generate face images of real people that are essentially undetectable as synthetic to most human observers. Journal Article Cognitive Research: Principles and Implications 10 1 Springer Science and Business Media LLC 2365-7464 Face perception; Deepfakes; ChatGPT; Artificial intelligence 14 10 2025 2025-10-14 10.1186/s41235-025-00683-w COLLEGE NANME Psychology School COLLEGE CODE PSYS Swansea University Another institution paid the OA fee This work was supported by the Israel Science Foundation grant (ISF-1498/21) to DF. 2025-11-14T12:50:22.9035411 2025-10-15T20:20:29.5972451 Faculty of Medicine, Health and Life Sciences School of Psychology Robin S. S. Kramer 0000-0001-8339-8832 1 Alex Jones 0000-0003-3600-3644 2 Daniel Fitousi 3 Jeremy Tree 0000-0001-6000-8125 4 70692__35634__eb5820ee81f848e296a3a357e1d11381.pdf 70692.VoR.pdf 2025-11-14T12:47:42.6710431 Output 1910872 application/pdf Version of Record true © The Author(s) 2025. This article is licensed under a Creative Commons Attribution 4.0 International License. true eng http://creativecommons.org/licenses/by/4.0/
title AI-generated images of familiar faces are indistinguishable from real photographs
spellingShingle AI-generated images of familiar faces are indistinguishable from real photographs
Alex Jones
Jeremy Tree
title_short AI-generated images of familiar faces are indistinguishable from real photographs
title_full AI-generated images of familiar faces are indistinguishable from real photographs
title_fullStr AI-generated images of familiar faces are indistinguishable from real photographs
title_full_unstemmed AI-generated images of familiar faces are indistinguishable from real photographs
title_sort AI-generated images of familiar faces are indistinguishable from real photographs
author_id_str_mv a24e1e2a89b0a9120fe03b481a629edd
373fd575114a743d502a979c6161b1ad
author_id_fullname_str_mv a24e1e2a89b0a9120fe03b481a629edd_***_Alex Jones
373fd575114a743d502a979c6161b1ad_***_Jeremy Tree
author Alex Jones
Jeremy Tree
author2 Robin S. S. Kramer
Alex Jones
Daniel Fitousi
Jeremy Tree
format Journal article
container_title Cognitive Research: Principles and Implications
container_volume 10
container_issue 1
publishDate 2025
institution Swansea University
issn 2365-7464
doi_str_mv 10.1186/s41235-025-00683-w
publisher Springer Science and Business Media LLC
college_str Faculty of Medicine, Health and Life Sciences
hierarchytype
hierarchy_top_id facultyofmedicinehealthandlifesciences
hierarchy_top_title Faculty of Medicine, Health and Life Sciences
hierarchy_parent_id facultyofmedicinehealthandlifesciences
hierarchy_parent_title Faculty of Medicine, Health and Life Sciences
department_str School of Psychology{{{_:::_}}}Faculty of Medicine, Health and Life Sciences{{{_:::_}}}School of Psychology
document_store_str 1
active_str 0
description Human users are now able to generate synthetic face images with artificial intelligence (AI) tools. Although indistinguishable from real photographs, these images have tended to feature fictional identities that do not exist in the real world. As a result, their use in applied contexts, including the spread of fake information, is similarly limited. Here, we investigated a new method for generating face images (via ChatGPT plus DALL·E) and its application to both fictional and real (in this case, celebrity) identities. Our results demonstrated that generated images of both fictional (Experiment 1) and celebrity identities (Experiment 2) could not be distinguished from real photographs. Further, providing additional real photographs for comparison during the task resulted in limited gains (Experiments 3 and 4). Finally, prior familiarity with celebrity faces produced only modest performance improvements. Therefore, new methods of detection should be explored as a matter of urgency since the latest ‘off the shelf’ AI tools can now generate face images of real people that are essentially undetectable as synthetic to most human observers.
published_date 2025-10-14T18:10:04Z
_version_ 1850692814909734912
score 11.08899

Similar Items