Creator-class collective bargaining and litigation on generative-AI training and likeness

Writers, actors, illustrators, voice artists, and authors use the institutions of their professions — guilds, unions, professional associations, trade groups — to negotiate or litigate protections against the unconsented use of their work and likeness to train and deploy generative AI. The instruments are familiar: industry contracts (the WGA / SAG-AFTRA settlements), copyright class actions against model developers, federal-policy advocacy on training-data disclosure, professional-association codes of conduct.

An actor chooses this because creators have an unusually well-developed pre-existing organising substrate for an industrial demand and a clean legal anchor (copyright, right of publicity) that AI training plausibly violates. The fight gives the broader make-AI-good movement what it otherwise lacks: a constituency of named, sympathetic individuals, an immediate economic case the press can carry, and a body of case law that may set the consent-and-compensation default for everyone else's data, not just creators'.

It trades off generality for traction. Wins flow first to the unionised tier — Hollywood writers and actors, big-press authors — and risk leaving non-unionised creators (most illustrators, most freelance journalists) under worse terms after the settlement than before. And copyright is a narrow weapon for what is, structurally, a labour and consent question.