The prevailing response to AI-generated disinformation — watermarking, C2PA provenance tracking, deepfake detection classifiers — rests on a premise Iran's campaign has already invalidated: that influence operations require audiences to mistake synthetic content for authentic documentation. The is not the fabricated news clip that needs to pass forensic review but the meme that travels because it entertains. Renée DiResta's assessment in Time that points to the structural gap: policy conversations about AI and disinformation consistently treat distribution as a…