Hollywood stars unite behind new AI consent standard

Hollywood is taking a tech-first approach to stop AI clones.

RSL Media, a nonprofit cofounded by Cate Blanchett, just launched the Human Consent Standard. It is a new licensing system backed by an A-list coalition including George Clooney, Tom Hanks, Meryl Streep, Viola Davis, Kristen Stewart, and director Steven Soderbergh, alongside major players like the Creative Artists Agency (CAA) and the Music Artists Coalition.

Teaching robots.txt to respect human likeness

Until now, protecting your likeness from AI scrapers meant playing trademark whack-a-mole. Matthew McConaughey trademarked video clips of himself, while Taylor Swift filed trademarks for soundbites like "Hey, it’s Taylor."

The Human Consent Standard takes a very different route by plugging directly into existing web infrastructure. AI crawlers can discover these terms through a site's standard robots.txt file, the basic gateway file crawlers check before scraping.

RSL Media cofounder Eckart Walther noted that while previous web standards applied only to specific URLs, this standard protects the underlying identity, work, character, or voice wherever it appears on the web.

In June, the group will launch a public verification registry. Users will be able to confirm their identity and choose between three clear options: grant AI systems full access, allow conditional or paid access, or restrict usage entirely. RSL Media then translates these preferences into machine-readable signals.

Blanchett stressed that the tool is designed to be free and accessible to everyday creators, not just Hollywood royalty.

Why it matters

This is a clever attempt to standardize AI licensing before governments step in with clunky legislation. Instead of demanding a whole new internet protocol, RSL Media is hacking robots.txt into a universal copyright badge for the human face and voice.

For creators and actors, it provides a unified front. But there is an obvious vulnerability: robots.txt is fundamentally an honor system. Rogue scrapers and smaller model builders routinely ignore crawler rules with zero consequence.

We think the true power of this registry is legal leverage rather than technical blockade. By putting machine-readable "do not scrape" flags on record, rights holders eliminate the defense of fair-use ambiguity. If an AI lab scrapes an actor who logged a conditional-access flag in the registry, it turns an accidental training scrape into willful infringement.

The real test starts in June. The standard will only matter if the leading frontier AI labs agree to bake the registry into their data ingestion pipelines.