As artificial intelligence advances and proliferates, outputs generated via neural networks—text, music, and visual works—increasingly resemble human creations, challenging traditional copyright frameworks. Although formally work-like, such outputs depend on human intellectual input and technological guidance, and should be regarded as extensions of human labor. To address ambiguities in legal characterization, ownership, and authorship of AI-generated content, this paper proposes a sui generis copyright framework emphasizing clear criteria, user centered ownership, and refined attribution rules.