Grok AI deepfakes of world leaders mislead social media users
Multiple deepfake images of world leaders including Javier Milei, Donald Trump, Giorgia Meloni, and Santiago Abascal were created with Grok, X's AI image generation tool, and shared as authentic photos on social media in January 2025. Fact-checkers confirmed the images were AI-generated, though they had already circulated widely and gained significant engagement before being debunked.
What happened
Multiple AI-generated images of world leaders were created using Grok, X's generative AI image tool, and circulated on social media as authentic photographs during January 2025.
One image showed Argentine president Javier Milei kissing the US flag while Donald Trump applauded. Another image depicted Milei alongside Trump, Italian prime minister Giorgia Meloni, and Spanish Vox party leader Santiago Abascal. Both images bore Grok's watermark.
Reported impact
The deepfakes circulated widely and were shared as real by users, including Colombian president Gustavo Petro, who later deleted his post sharing the Milei flag-kissing image. The Milei-Trump-Meloni-Abascal image received over 25,000 likes on X before being identified as AI-generated.
Index analysis
Fact-checking organizations Maldita.es, Chequeado, VerificaRTVE, and Hive Moderation identified the images as artificially generated. Hive Moderation determined the group image had a 97.6% probability of being AI-generated. The images bore Grok's visible watermark.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.