Abstract
<p>The municipal elections of March 15 and 22, 2026 constitute the firstFrench local election in which generative artificial intelligence hasmassively entered campaign practices. This research brief offers abalanced analysis of documented uses, distinguishing legitimateapplications (poster generation, programmatic chatbots, contentsynthesis) from confirmed abuses (defamatory deepfakes, misleadingimages, astroturfing). While these new technological capabilities raiselegitimate concerns for the integrity of democratic debate, we argue thattheir impact must be put in perspective in light of the literature on thelimited effects of mass political communication and the actualdeterminants of disinformation consumption. Rather than acatastrophism that risks fueling generalized distrust, we advocate for anapproach grounded in media literacy and the strengthening ofprofessional journalism.</p>