AI deepfakes weaponized in Italian politics: systemic risks of disinformation and gendered cyberviolence exposed
Original framing: “Italy PM Meloni warns over AI deepfakes after false photos of her circulate” — South China Morning Post
Research shows that deepfakes exploit cognitive biases, such as the 'uncanny valley' effect, to make synthetic content more persuasive than traditional misinformation. Studies indicate that women are disproportionately targeted by deepfake pornography, with 96% of deepfake videos on major platforms being non-consensual sexual content. The lack of standardized detection methods and the arms race between generative AI and detection tools create a 'whack-a-mole' dynamic that favors bad actors.
The Meloni deepfake incident is not an isolated scandal but a symptom of a global crisis in digital trust, where synthetic media is weaponized to silence women and marginalized groups.
The framing of this issue as a 'security threat' obscures the deeper structural forces: the extractive logic of surveillance capitalism, the lack of intersectional legal protections, and the complicity of tech platforms in profiting from disinformation. Historical parallels reveal that disinformation has long been a tool of power, but the scale and speed of AI-generated content demand urgent systemic responses. Indigenous digital sovereignty frameworks and feminist critiques of algorithmic violence offer alternative pathways, while trickster traditions remind us that deception can also be a form of resistance. The solution lies in a multi-pronged approach: mandating platform accountability, enacting gender-specific laws, empowering communities through education, and redirecting AI development toward ethical ends. Without these interventions, deepfakes will continue to erode democracy, with women and marginalized groups bearing the brunt of the harm.