Robin Williams' daughter pleads for people to stop sending AI videos of her dad (www.bbc.co.uk)

Zelda Williams, daughter of late actor Robin Williams, publicly urged people to stop sending her AI-generated videos of her father, calling them hurtful, exploitative and “not what he’d want.” Her plea highlights a growing social-media trend that animates images and reconstructs voices of deceased people—often with minimal fidelity—framed as “bringing loved ones back.” Williams has previously described voice-cloning attempts as “personally disturbing” and criticized these works as crude re‑assemblies of real lives and artistic legacies for cheap engagement. The episode underscores broader technical and ethical issues in generative AI: deepfake video and voice-cloning models are trained on large datasets of real performers’ work, producing outputs that may “vaguely” resemble individuals without consent or context. Unions like SAG‑AFTRA have pushed back—saying synthetic “AI actors” (e.g., Tilly Norwood) are built from the labor of countless professionals and lack lived experience or emotion—raising questions about dataset provenance, intellectual property, consent, and labor displacement. For the AI/ML community, this amplifies the need for robust guardrails: provenance metadata, opt-outs for posthumous likenesses, clear consent mechanisms, and industry standards to balance creative experimentation with respect for individuals and artists’ rights.