Indian courts are increasingly confronting cases where celebrities invoke personality rights — rooted in Articles 19 and 21 — to restrain AI-generated deepfakes, unauthorised commercial use of their name, image, or voice. The legal challenge is not to privilege one right over the other but to ensure that protecting personality rights does not extinguish legitimate satire, parody, criticism, and artistic expression that free speech under Article 19(1)(a) guarantees. Courts must calibrate proportionality, not pick sides.

Personality rights in India are not codified in a single statute. They emerge from the intersection of two constitutional provisions: Article 21, which protects the right to privacy and dignity (traced through Puttaswamy v. Union of India, 2017), and Article 19(1)(a), which ironically also protects the very expression that celebrities seek to restrain.

The doctrine has two components. The right of publicity protects a person's commercial interest in their identity — name, likeness, voice, and persona — from being exploited without consent, particularly for commercial gain. The right of privacy protects the personal, non-commercial dimension of identity from unwanted intrusion or misrepresentation.

The tension arises because Article 19(1)(a) protects not just news reporting but satire, parody, fan art, criticism, and artistic reimagination. Courts have consistently held that free speech can only be curtailed under the reasonable restrictions in Article 19(2), which do not include 'protecting a celebrity's brand value.'

The proportionality test — adopted from the Puttaswamy nine-judge bench — is the operative standard. Any restriction on expression to protect personality rights must be: (i) backed by law, (ii) pursuing a legitimate aim, (iii) necessary, and (iv) proportionate to the harm. A blanket injunction against all AI-generated content featuring a celebrity's likeness would fail this test. A targeted injunction against a deepfake advertisement misrepresenting commercial endorsement would likely pass it.

The deepfake dimension has sharpened the debate. Unlike traditional misrepresentation, AI-generated content can fabricate consent and create entirely fictional but convincing versions of real persons. Courts now grapple with whether existing IP and tort law — passing off, defamation — is adequate, or whether a statutory personality rights framework is needed.