AI systems devour voices, faces and performances at scale. They generate new content that sounds or looks identical to the originals. Yet the people whose identities fuel this process often receive nothing. Now a group of prominent actors aims to change that equation with a technical specification designed to make consent visible to machines.
The effort comes from RSL Media, a public benefit nonprofit launched on May 12, 2026. It extends an existing licensing protocol called Really Simple Licensing, or RSL, into the domain of human identity. The draft RSL Media Human Consent Standard (RSL-MEDIA) 1.0 lets individuals declare whether AI systems may train on or generate content from their name, voice, likeness or other attributes. Those declarations sit in a public registry scheduled to open next month.
Cate Blanchett, an Oscar-winning actress and co-founder of RSL Media, put the stakes plainly. “AI technologies are expanding rampantly, essentially unchecked and unregulated. In order for humans to remain in front of these technologies, consent must be the first consideration.” She described the project as “a simple, effective and free solutions-based technology for facilitating and activating consent.” The standard, she added, offers “the industry’s first practical solution where people everywhere, not just public figures, can assert control over how their work is used by AI.”
Blanchett’s words appear in the organization’s launch announcement. They carry weight in an industry where synthetic performances already threaten traditional work. Voice actors have watched their craft shift. Some now license clones for higher fees while others face replacement without payment. A report from Rest of World in April detailed how cloned voices handle dubbing indefinitely once captured, raising concerns about consent and ongoing compensation.
RSL Media builds directly on RSL 1.0, which launched in December 2025. That earlier standard gave publishers a machine-readable way to set licensing terms for web content scraped by AI. More than 1,500 organizations now back it, according to the RSL Collective. Sites including The Guardian publish RSL XML files that crawlers can read. The new media-focused version adds four categories of protected subjects: creative works, personal identity, characters and marks.
The technical design is straightforward yet precise. Rights holders reserve a unique International Standard Rights Declaration identifier. They publish XML documents containing
Nikki Hexum, co-founder and CEO of RSL Media, emphasized accessibility. “AI can’t respect rights it can’t see, and this means human consent is virtually invisible in this new digital era. The right to decide whether AI can use your work or identity should not be reserved for only those who can afford lawyers or have platforms big enough to be heard, it is a basic human right.” The registry aims to give ordinary people the same tools once limited to studios or agencies.
Support stretches across Hollywood. Dame Emma Thompson called the initiative “urgent and essential.” She noted that “AI is merely stealing from us all” but added that the solution is “eminently doable, so let’s do it without delay.” Director Steven Soderbergh praised its simplicity and resistance to manipulation. Dame Helen Mirren drew a sharp line between inspiration and imitation, labeling the latter “crass theft.” CAA co-chairman Kevin Huvane and the Music Artists Coalition also endorsed the project.
Yet questions linger about enforcement. The Register’s Thomas Claburn observed that legal consequences for ignoring such registries remain unclear. Data brokers and deepfake distributors have faced limited repercussions despite public outrage. AI companies have largely trained models on vast datasets first and addressed complaints later. More than a hundred copyright lawsuits target training practices, but few have produced definitive rulings.
This new standard joins a crowded field. Alternatives include TDMRep from the European EDRLab, Spawning’s ai.txt files, and IETF work on AI preferences. Cloudflare offers commercial pay-per-crawl options. RSL Media’s advantage lies in celebrity backing and its focus on identity alongside content. The specification explicitly protects minors by ignoring any permissions involving their likenesses.
Recent union gains add context. SAG-AFTRA secured expanded AI protections in a tentative four-year studio deal reached in early May 2026. The agreement builds on 2023 terms that required consent and compensation for digital replicas. It adds tighter rules on synthetic performers. The guild has also pursued legislation such as the NO FAKES Act, which would create a federal right over voice and visual likeness.
But contracts and laws move slowly. Technical standards promise real-time signals at internet scale. James Everingham, co-author of the RSL Media standard and former Instagram engineering lead, described the problem of fragmented rights scattered across contracts and databases. “RSL Media provides a critical infrastructure layer by translating consent and usage rights into a format that can work across systems, for both individuals and AI.”
The registry launch in June will mark the first test. Users will verify identity, declare permissions across the four subject areas, and generate machine-readable files. AI developers can then check before training or generating. Payment mechanisms may follow for approved uses. Success depends on adoption by both creators and the platforms that build the models.
Critics might dismiss another specification as pointless in a landscape already littered with them. Yet the involvement of A-list talent changes the dynamic. When Cate Blanchett, Emma Thompson and Helen Mirren speak on consent, studios and tech firms listen. Their participation could push RSL Media beyond theoretical protocol into practical industry norm.
And the timing feels urgent. Voice cloning tools from companies like ElevenLabs now produce convincing results from short audio samples. Pricing comparisons show commercial services charging fractions of a cent per generated minute. The economic incentive to replace rather than hire grows stronger each quarter. Performers risk losing not only current gigs but future residuals as synthetic versions persist indefinitely.
Jacqueline Sabec, a partner at King, Holmes, Paterno & Soriano and co-author of the standard, stressed the need to protect human creativity while enabling responsible innovation. Francesca Amfitheatrof, a jewelry designer and former Tiffany design director, framed creativity as “the most precious and irreplaceable expression of what makes us human.” Their perspectives broaden the coalition beyond pure entertainment.
RSL Media positions itself as infrastructure rather than advocacy. It offers a traffic-light system: green for allowed uses, red for prohibited, yellow for conditional with payment or attribution. The XML format integrates with existing web mechanisms such as sitemaps. Discovery happens through predictable URLs or embedded metadata. The design draws from decades of web standards experience, including Robots Exclusion Protocol, which many AI crawlers have ignored.
Whether this effort succeeds where others faltered will depend on implementation details still unfolding. The specification remains a draft. Registry policies, dispute resolution and integration with payment systems need definition. Yet the core idea is clear. Make consent machine-readable. Give every person, famous or not, a practical way to set terms. Force AI systems to see the humans behind the data.
Blanchett and her colleagues have bet that Hollywood muscle combined with clean technical design can shift the balance. The coming months will reveal if the machines, and the companies that build them, choose to look.
WebProNews is an iEntry Publication