Opinion: Oscar rules drew a line on AI performances—platform safety is still the harder problem

The Academy’s May 2026 eligibility updates do something awards bodies rarely do cleanly: they state, in plain English, that fully synthetic performances cannot compete for acting Oscars and fully synthetic screenplays cannot compete for writing Oscars. Consent and legal billing matter—not vibes alone.

It is equally important to say what the rules do not fix. They do not ban generative tools from marketing, localization, concept art, or many visual-effects pipelines where branch juries weigh how much human judgment shaped the final work. They do not stop likeness theft, voice cloning, or credit fights in writers’ rooms. Oscar policy is a narrow screwdriver; labour law, healthcare access, and platform regulation are whole toolboxes.

Separately, serious reporting has documented harms tied to immersive chat products—cases where vulnerable users are steered into harmful narratives. That is a different regulatory problem than eligibility for a statue, but it sits on the same societal axis: who answers when scalable persuasion meets loneliness or mental illness?

Durable trust requires dull systems: consent logs for synthetic likeness, fast takedowns when deepfakes hijack a campaign, protections for juniors asked to “human-wash” machine drafts, and union contracts that keep pace with model releases. Awards-season rules cannot replace HR, security, and clinical capacity.

Globally, personality rights and platform liability differ by country. A film’s compliance story in California is not identical to its risk story in London, Seoul, or São Paulo.

None of this diminishes the Academy’s step; it places it. Clear lines help audiences know what a trophy means. They do not, alone, protect teenagers from predatory prompts at two in the morning.

This column represents the Newsorga editorial board. Interpretive claims are ours; factual statements about Academy rules should be checked against the Academy’s own publications.