Hollywood stars back a standard for controlling how AI uses faces and creative work

Hollywood stars and directors are backing a new licensing standard for artificial intelligence systems. The Human Consent Standard lets people set rules for how AI systems use their face, voice, characters and creative work.

How the standard works

The Human Consent Standard gives people three ways to control their content and likeness. It is simple and direct: you can give AI full access to your material, set special terms of use (for example, requiring a license agreement and payments), or block access completely, with no exceptions.

Each option works as a signal to AI systems. It is enough to add one special file to a website or profile, and systems will know your preferences. This is similar to how robots.txt tells search engines which pages to index, except that here the signal is aimed at generative AI and models trained on content. Ideally, large AI developers will check this file before they start training their models.

• Full permission: AI can use the content freely, without restrictions
• Conditional access: a license and payment are required for commercial use
• Full ban: AI cannot use the content for any purpose

Who launched the standard

The Human Consent Standard was developed by RSL Media, a nonprofit organization that created the base Really Simple Licensing (RSL) Standard last year. The people behind it are not just random directors and actors but global stars: George Clooney, Tom Hanks and Meryl Streep. Their support is a strong signal from the Hollywood establishment that the use of creative work in the AI era needs to be controlled.

The original RSL Standard was launched as a way for websites and authors to signal to AI systems how they want their content to be used. Now the standard is expanding to a new level: protecting a person's likeness and creative output. This matters especially for actors, who can be synthesized into video without their consent, and for musicians, whose voices can be cloned.

Why this matters right now

Without such a standard, creative people remain defenseless against companies that train AI on everything, without asking, without paying and without credit. This has already affected many actors and musicians whose faces and voices appear in synthesized videos. Litigation is expensive and slow, and results often arrive too late, after the harm is done.

The Human Consent Standard offers a different path: simply put the licensing standard in place, and it works automatically. Platforms and AI developers will see your choice and respect it, if they want to stay on good terms with the industry and avoid lawsuits.

If large platforms and generative AI systems start respecting the standard, it may become the de facto standard for the whole industry. That would reduce conflicts between content creators and the companies that build AI. It is not yet clear whether companies such as OpenAI or Meta* will follow such a standard voluntarily.

What this means

The new standard is an attempt to set the rules of the game at the technology level, without government involvement. It shows that the industry is ready to seek a compromise and set its own rules rather than wait for court rulings and regulation. For authors, actors and musicians, it is a tool for protecting their rights. For AI companies, it is a way to be on the right side of history.

*Meta is designated an extremist organization and banned in Russia.