Artificial Elections 2.0: Generative AI in the 2026 French Elections

AI Forensics extended its earlier methodology from the 2024 study to examine official Facebook, Instagram, TikTok, X, and YouTube accounts of parties competing in France's 2026 municipal elections, observing content from January 14 to March 14, 2026. Unlike the prior study, this round focused specifically on parties running in the municipal races rather than the full national political landscape.

The research found that most AI-generated campaign content went unlabelled even when technical watermarks such as SynthID were present in the file, indicating parties and platforms were not surfacing existing disclosure signals to viewers. Generative AI use expanded beyond photorealistic imagery into stylized and cartoonish visuals used to reinforce campaign messaging, and its use was increasingly normalized as part of standard campaign production rather than an exception. Usage concentrated heavily on right-wing and far-right parties, particularly Renaissance and Reconquête, with anti-immigrant narratives a recurring theme; one instance was also attributed to the Communist Party.

The report concludes that clearer enforcement of AI-content labeling under the Digital Services Act is needed to address the growing electoral use of unlabelled generative media.