On July 10, 2026, Meta announced the cessation of the Muse Image tool on Instagram, which allowed users to generate AI images by @mentioning public accounts. The feature was launched around July 7 along with the Muse Image model introduced by Meta Superintelligence Labs, with the system defaulting to include photos from public adult accounts as reference material, excluding accounts under 18 and private accounts.
The feature's mechanism was based on default participation: after a user tagged a public Instagram account, the platform automatically extracted its public photos as AI generation material without prior notification to the account holder. Meta initially stated that controls were in place, but in practice, users had to actively opt out to avoid being referenced. This design shifted the consent responsibility to individuals, rather than requiring the platform to obtain explicit permission before generation.
For the platform itself, this withdrawal means that other AI effects in Instagram Stories can continue in the short term, but the planned expansion to integrate into WhatsApp and the standalone Meta AI app has been hindered. As a result, Meta Superintelligence Labs' first model is restricted in its core social referencing capabilities, affecting its pace in catching up with other image generation services.
Creators and the entertainment industry face direct risks regarding portrait rights. SAG-AFTRA advised its members to disable the feature and welcomed Meta's withdrawal, pointing out that the platform's default mechanism misjudged the public backlash against involuntary digital replication. CAA emphasized that names, likenesses, and creative works must not be used by third parties without clear documented consent, and protecting creators' livelihoods must be a prerequisite for innovation.
Privacy advocacy organizations pointed out that such features treat public images as raw materials ready for exploitation. The London-based international privacy organization told the BBC that this once again demonstrates AI companies' approach to processing personal data. Apar Gupta, founding director of the Internet Freedom Foundation, criticized Meta for long-term privacy violations under the guise of social platform operations.
Cross-references show that similar default mechanisms have sparked backlash in other services. After OpenAI launched Sora 2, it adjusted due to the ability to generate content with trademarked characters, and later reached a partnership with Disney, but Sora 2 has ceased operations. Meta's quick halt this time reflects a common pattern of platforms adjusting strategies under pressure from creator rights.
On the regulatory front, California has passed AB 2602 and AB 1836, requiring permission before commercially using AI-generated images of living or deceased performers. The EU GDPR considers images of individuals as personal data, requiring a legal basis under Article 6 and meeting the requirements of free, specific, informed consent under Article 7. The EU AI Act further requires informing the public about AI-generated or AI-enhanced content.
For developers, future AI image features must establish a prominent prior consent process before launch, rather than relying on opt-out options. Otherwise, they will face joint opposition from unions, talent agencies, and regulators, increasing compliance costs and delaying product iteration.
The gains and losses for enterprise users and ordinary creators lie in the return of control: public content is no longer automatically used as generation material, but platforms may shift to stricter review or paid licensing models, reducing the space for free experimentation.
© 2026 Winzheng.com 赢政天下 | 转载请注明来源并附原文链接