It’s Sora again.
Meta, led by CEO and AI enthusiast Mark Zuckerberg, this week unveiled Muse Photo and Muse Video, the first photo and video tools born from its superintelligence lab.
The tool puts Meta on the cutting edge of AI media creation alongside Google’s Gemini and Seedance 2.0, but along with the tool, Meta also said that all public Instagram profiles will be opt-out rather than opt-in, and that other users can bring in other accounts that haven’t opted out and use their profiles to create new content.
For celebrities and other public figures, it can be a big problem, even if most eventually opt out. Now, the Creative Artists Agency is calling on Meta to make the move, asking the tech giant to make it an opt-in rather than an opt-out.
“An individual’s name, image, likeness, voice, or creative work may not be used by any third party, including AI models, without explicit written consent. True innovation is about putting creators first, respecting their rights, protecting their lives, and giving them real control, not relegating them to a platform,” CAA said in a statement Wednesday night. “We have expressed our concerns to Meta on behalf of our clients, expressing our disapproval and view of the need for a more responsible approach. We are asking Meta to make Muse Image protection the default, not the exception, and to allow individuals to opt-in if they wish to allow the use of their images and likenesses for AI content creation.”
“Artists have the right to decide whether and how their likeness and work will be used, based on consent and the ability to set their own terms,” the statement continues. “This means creators can impose limits, monitor usage, and prevent fraudulent recommendations and abuse. Responsible AI requires clear disclosure and prompt removal of abusive content. There must be easy ways to discover, track, and remove abuse, and it must be obvious that something was generated by AI. CAA We believe in the power of new technology, but not at the expense of individual rights and livelihoods. The future of creativity depends on respecting the ownership and autonomy of those who create it.”
Of course, CAA is well aware of the opportunities and threats that generative AI poses to its clients. The agency maintains what it calls the “CAA Vault,” where clients’ digital likenesses are stored for future monetization and protection, and the agency was also the first test user of YouTube’s deepfake detection tool, which was rolled out across Hollywood earlier this year.
Muse’s situation is reminiscent of OpenAI’s failed Sora launch, which was filled with big-name IP and many celebrity likenesses, forcing the AI giant to pivot to an opt-in model before ditching the video model altogether earlier this year.
After Sora’s launch, WME announced it would opt out on behalf of all its customers, the first domino to fall in the model’s short life.
“As artists and creatives encounter AI models that leverage their intellectual property, not just their names, images and likenesses, there is a strong need for real protection,” Chris Jacmin, head of digital at WME, wrote in a memo at the time. “Our position is that artists should have a choice in how they appear in the world and how their likeness is used, and we have informed OpenAI that all WME clients should opt out of the latest Sora AI update, regardless of whether the intellectual property owner has opted out of the IP associated with our client.”
With so many celebrities, artists, creators, and other public figures on Instagram, an opt-out model may not be enough, especially if their likeness is what they make a living from.
