Skip to content
Submit incident
AII-1002-02309
ReportedReported by fewer than three independent publishers, with no official finding.Supporting sources are pending verification.

AI-Generated Videos Falsely Depict Protests Against Brazilian Politician

AI-generated videos using Google's Veo tool circulated on TikTok depicting fabricated protests against Brazilian Chamber President Hugo Motta, with watermarks visible but no clear warnings of synthetic origin. Verification by Comprova confirmed the videos were AI-created, though some social media users mistook them for genuine demonstrations.

4 July 2025 Β· Google Β· Misinformation and synthetic media
4 July 2025
AII-1002-02309
Quick facts
Category
Misinformation and synthetic media
Country
Brazil
Date
4 July 2025
Drafted by
Raidu Incident Detection

What happened

Videos circulating on TikTok depicted hundreds of people in streets protesting against the Brazilian National Congress and Chamber of Deputies President Hugo Motta (Republicanos-PB), with signs reading 'Congress enemy of the people' and chants of 'Hugo Motta, traitor'.

The videos bore watermarks identifying them as generated by Veo, Google's AI video generation tool. Although marked as AI-generated, the videos lacked clear warnings of synthetic origin, leading some users to believe they documented real protests.

Reported impact

Comments on the posts indicated users believed the videos represented genuine protests, with one stating 'Wow, I got goosebumps, it is well said that the voice of the people is the voice of God' and another claiming 'The people seeing REALITY'.

Index analysis

Verification by Comprova using Google's SynthID Detector, a tool recognizing AI-generated content from Google products, indicated the videos were created with AI with a high degree of probability.

Classification

Category
Misinformation and synthetic media
Industry
Media, social platforms and marketing
Country
Brazil
Harm status
Actual harm
Intent
Unknown
Timing
Post-deployment
Harm types
Public interest, Reputational
Harmed
General public
Failure modes
Deepfake video, Misidentification
AI systems
Veo

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident