False AI images circulate around French protests
Fact-checkers found generated images and old footage shared as current violence in France.
Listen to this article
There is no audio version yet. Request one and an AI voice will read the article aloud.
Read by an AI voice.
AI-generated images and old videos are circulating during protests in France, presented as current events. Spanish and British fact-checkers exposed several examples, while genuine violence and looting were also recorded.
VerificaRTVE examined images of a person dressed as the Pink Panther among demonstrators. The images carried a label indicating that they had been generated with artificial intelligence; analysis using SynthID and TruthScan indicated that the photo and video had been created or edited with AI.
A video of a burning building was also incorrectly labelled. According to RTVE, the footage was not recorded in the French city of Limoges, but in Tbilisi, Georgia, during protests in 2024. The video was redistributed as if it showed the recent violence around Limoges town hall.
Another widely shared image was said to show demonstrators with Palestinian flags near the Eiffel Tower. Full Fact traced the image to a satirical article from October and assessed the claim as manipulated. This does not mean that all images of the protests are false.
That combination is precisely what makes verification difficult. Demonstrations, clashes and damage have genuinely been reported in France in recent days. A real news event can therefore provide the backdrop for old, manipulated or entirely fabricated images.
The examples also show the limits of automatic detection. A platform label or analysis by a fact-checker provides indications, but investigating provenance, comparing images and checking place and time remain necessary. For readers, the main message is simple: do not share an image before checking its original source and date.
Also show where in Europe and in the world
One story, several perspectives
What is established
- Several misleading images were circulated during French protests.
- Fact-checkers identified AI-generated images and old images from other countries.
- There were also genuine protests and incidents in France.
Left
Arguments Platforms must take greater responsibility for misleading images that set social groups against one another. Transparency about recommendation systems and swift corrections matter more than maximum freedom of distribution.
Values Protecting public information, inclusion and social safety.
Consequences Stricter moderation can limit harmful campaigns, but must be accompanied by appeals processes so that legitimate images are not suppressed.
Centre
Arguments The best approach combines platform responsibility with independent journalism, media literacy and clear sourcing. Not every incorrect image requires removal; context and proportionality are decisive.
Values Reliability, freedom of expression and institutional oversight.
Consequences More verification takes time, but reduces the risk that a truthful report will be overshadowed by false images.
Right
Arguments The spread of false images shows that citizens and national governments must protect their own information provision. Platforms must not decide what is true without democratic oversight.
Values National control, order and freedom of expression.
Consequences Greater emphasis on source transparency and criminalising deliberate manipulation may strengthen trust, but broad censorship could affect political criticism.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The examples described are based directly on fact-checks by RTVE and Full Fact. The text avoids an unsubstantiated claim about the total volume of disinformation.
- confirmed AI-generated images of the Pink Panther at French protests were circulated online. — RTVE describes the circulation and the technical analysis. source
- confirmed A video shared as a fire in Limoges came from Georgia in 2024. — RTVE traced the video through reverse searching and geolocation. source
- confirmed An image of demonstrators near the Eiffel Tower was AI-generated. — Full Fact traced the image to a satirical source and assessed it as manipulated. source
Editor's note
The specific examples were checked by RTVE and Full Fact. The scale of their circulation and their impact on public debate cannot be measured fully.Sources
- Bulos y desinformación sobre las protestas en Francia — RTVE Verifica
- Viral image of protesters waving Palestinian flags outside the Eiffel Tower is AI-generated — Full Fact
- SynthID — Google DeepMind
More on this in Dutch media
- de Volkskrant — „desinformatie kunstmatige intelligentie”
- RTL Nieuws — „desinformatie kunstmatige intelligentie”
- NOS — „desinformatie kunstmatige intelligentie”