Meta has suspended its newly launched Muse Image feature, which allowed users to generate AI images by tagging public Instagram accounts, after facing swift backlash from Hollywood talent agencies and performers' unions over its consent model. The company introduced the tool on Tuesday but disabled it by Friday, marking another high-profile retreat in the ongoing tension between AI companies and creative professionals over how likenesses and intellectual property are used in generative models.

What Went Wrong With Meta's Approach to Creator Consent?

The core issue centered on Meta's "opt-out" policy, which required users over 18 with public Instagram accounts to actively disable the feature if they didn't want their images used in the AI generator. This meant that by default, anyone's public profile could be referenced to create AI-generated content without their explicit permission. Creative Artists Agency (CAA), which represents major stars including Tom Hanks and Meryl Streep, immediately objected to this approach.

"No one's name, image, likeness, voice or creative work should be used by any third party, including AI models, without clear, documented consent," CAA stated in a message to Meta.

Creative Artists Agency, statement

SAG-AFTRA, the union representing U.S. performers, also urged its members to opt out of the feature, posting on social media to "take action to protect your likeness". The rapid mobilization from both talent representation and labor organizations signaled that the industry had learned lessons from previous AI rollouts.

How Does This Compare to OpenAI's Sora Experience?

Meta's experience mirrors what happened with OpenAI's Sora video generation model, which faced similar criticism over creator consent. OpenAI initially launched Sora with an opt-out policy but quickly moved to an opt-in model after backlash. The situation escalated further when OpenAI announced it was shutting down Sora entirely earlier this year, ending a reported $1 billion partnership with Disney that would have allowed the entertainment company's intellectual property to appear on the platform.

Talent agency WME had taken proactive steps during Sora's rollout, opting all of its clients out of the model regardless of whether intellectual property rights holders had done so. At the time, WME's digital chief Chris Jacquemin explained the agency's position: "There is a strong need for real protections for artists and creatives as they encounter AI models using their intellectual property, as well as their name, image and likeness".

Chris Jacquemin

What Are the Key Differences Between Opt-Out and Opt-In Consent Models?

The distinction between these two approaches has become central to the debate over AI and creator rights. Understanding the difference helps explain why the industry reacted so strongly to Meta's initial policy.

• Opt-Out Model: Users must actively take steps to prevent their content from being used, meaning the default assumption is permission granted. This places the burden on individuals to protect themselves.
• Opt-In Model: Users must explicitly agree to allow their content to be used, meaning the default assumption is no permission unless stated otherwise. This places the burden on the AI company to request consent.
• Creator Control: Opt-in models give creators the ability to set their own terms, monitor usage, and prevent unauthorized endorsements or exploitation, which CAA emphasized as essential for protecting livelihoods.

CAA emphasized that "artists deserve to decide if and how their likeness and work is used, with consent and the ability to set their own terms". The agency also noted that there should be easy ways to spot, track, and take down misuse, and it should be clear when something is AI-generated.

How Are Talent Agencies Preparing for AI-Generated Content?

Beyond pushing back on individual features, talent agencies are taking structural steps to protect their clients' digital identities. CAA maintains what it calls the "CAA Vault," which stores the digital likenesses of its clients for future protection and monetization. The agency also became the first to test YouTube's deepfake detection tool before it was rolled out more widely across Hollywood, signaling that the industry is building infrastructure to manage AI-generated content at scale.

Meta acknowledged the feedback in its decision to suspend the feature. "Our intent was to provide a useful creative tool and to give people control over whether their public content could be referenced in this way," a Meta spokesperson said. "We've heard the feedback that this feature missed the mark, so it's no longer available".

Meta

After Meta's reversal, CAA commended the company's swift action, stating: "Putting individual rights and consent at the forefront is essential to building responsible technology. We look forward to ongoing conversations to ensure creators stay protected as technology evolves". The statement suggests that while the immediate crisis has passed, the broader conversation about how AI companies should handle creator consent is far from over.

Meta's decision to disable the Muse Image feature comes as Apple filed a federal lawsuit against OpenAI, alleging the ChatGPT maker stole company secrets with help from former Apple employees. OpenAI has denied the claims. The timing underscores how AI companies are facing mounting legal and reputational pressure as they expand their generative capabilities.