Flávio Bolsonaro AI media incident — Oct 2026
The campaign of Flávio Bolsonaro utilized AI-generated avatars and manipulated event photography to artificially inflate crowd sizes and simulate support from.
- Reported date
- Oct 21, 2026
- Target
- Flávio Bolsonaro
The exact incident date was not established. This entry is dated by its source report.
What happened
The presidential campaign of Flávio Bolsonaro employed a sophisticated digital strategy that integrated AI-generated content and automated networks to manufacture artificial public support. Central to this operation was the use of synthetic media to overcome demographic challenges and maintain a political presence for restricted figures. The campaign utilized AI image generation to alter event photography, specifically scaling up crowd sizes to exaggerate physical attendance and digitally swapping male attendees with female faces. This demographic engineering was designed to simulate broader support among female voters, a group where polling data indicated lower engagement for the candidate.
Beyond static image manipulation, the campaign deployed AI-generated digital avatars to broadcast endorsements from former president Jair Bolsonaro, effectively bypassing judicial rulings that restricted his direct participation in political activities. Furthermore, investigations identified over 50 TikTok videos featuring on-the-street interviews that were altered using AI to simulate support for Flávio Bolsonaro. While these videos garnered over 7 million views, only 18% included disclosures indicating the individuals featured were AI-generated.
These efforts were supported by industrial bot networks purchased from cybersecurity firms. These automated tools operated across WhatsApp, Instagram, and X to generate synchronized posts and artificial engagement. By deploying thousands of accounts programmed to post identical messaging scripts, the campaign artificially inflated engagement metrics and pushed specific political hashtags into regional trending lists. Technical evaluations submitted to electoral courts confirmed that these account clusters posted at frequencies impossible for human operators, creating a false impression of widespread popular momentum to sway undecided voters.