Abby Pownall
Volume 77, Issue 5, 1439-1466
Rapid advancements in artificial intelligence (AI) pose an emerging crisis for the spread of political misinformation. Generative AI, capable of producing hyper-realistic deepfakes, synthetic audio, and fabricated video, enables private individuals to construct and disseminate defamatory content at an unprecedented scale and speed. When libel against public officials is generated using these tools, the actual malice standard is satisfied as the user cannot produce such an output without knowingly and intentionally prompting the AI system to fabricate representations that appear factual. Though defendants may invoke the satire defense, the
increasing realism of deepfakes has blurred the line between satire and factual assertions. The psychological impact of images and videos exceeds that of text, rendering the reasonable person standard traditionally applied in cases of satire unstable, and ultimately calling into question courts’ ability to rely on that standard in cases involving synthetic media. This Note argues that the Supreme Court’s landmark decision in New York Times Co. v. Sullivan rests on empirical and psychological assumptions about how speech circulates, how reputational harm can be corrected, and how audiences distinguish fact from satire—assumptions that no longer hold in an information ecosystem dominated by generative AI and synthetic media. Sullivan remains a pillar of American free speech jurisprudence, and while its normative commitment to uninhibited debate remains sound, its doctrinal architecture must be recalibrated to account for how AI-generated content undermines the feasibility of counter-speech, the effectiveness of retractions, and the workability of the satire defense and the reasonable person doctrine.