Meta has abruptly removed a controversial artificial intelligence feature on Instagram that enabled users to generate and modify images by referencing public accounts, drawing sharp criticism from privacy advocates, talent agencies, and creators. The feature, part of the company's Muse Image tool developed by Meta Superintelligence Labs, was launched earlier this week but pulled within days after widespread backlash.
Muse Image was introduced as a creative tool that allowed individuals to generate new visuals by '@-mentioning' public Instagram accounts, effectively using those accounts as stylistic references or sources for AI manipulation. Notably, the feature did not inform account owners when their photos were used in this way, raising immediate red flags about consent and potential abuse. TechCrunch had published a guide on how to disable the feature, which itself became a testament to the widespread concern.
On Friday, Meta announced the removal in a blog post, saying: '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. We’ve heard the feedback that this feature missed the mark, so it’s no longer available.' The post, first shared by Puck News founding partner Dylan Byers, acknowledged that the company had not anticipated the level of controversy it would generate.
The Nature of the Backlash
The backlash was swift and multifaceted. Privacy experts argued that even though the accounts were public, using them as input for AI generation without explicit consent violated established norms of digital autonomy. Talent agencies, including Creative Artists Agency (CAA), voiced concerns that the feature could be exploited to create non-consensual explicit images, a problem that has long plagued AI image generators. The misuse of AI to generate sexually explicit deepfakes, particularly of female celebrities, has been a persistent issue across platforms. Instagram, with its vast library of public photos, was seen as a goldmine for such abuse.
Creators and influencers also criticized the feature for devaluing their work. Many earn income through their visual aesthetics, and allowing anyone to remix their style without attribution or compensation was seen as a direct threat. Meta had not built in any monetization or notification system, which fueled the perception that the company was prioritizing rapid AI deployment over ethical considerations.
Historical Context of AI Misuse on Social Media
This incident is part of a broader pattern of social media platforms launching AI features that later prove problematic. In 2023, Meta faced similar criticism for its 'Make-A-Scene' AI tool, which could generate images from text prompts but was also used to create abusive content. Earlier, in 2024, the company had to issue emergency patches for its 'Imagine' feature on WhatsApp after users discovered they could generate hateful images.
The misuse of AI in generating non-consensual imagery has been especially acute for female celebrities. In 2025, a wave of deepfake images of Taylor Swift went viral on Twitter (now X), prompting calls for stricter regulation. Instagram itself had previously introduced guardrails to prevent such abuse, but the rapid rollout of Muse Image showed that those protections were insufficient. The feature essentially allowed users to bypass existing safeguards by using public accounts as a reference source.
Implications for Privacy and Consent
The removal of Muse Image raises deeper questions about how social media platforms handle public data in the age of generative AI. While public photos on Instagram have long been viewable by anyone, they were never intended to be used as raw material for AI manipulation. The concept of 'implied consent' for public content has now been severely tested. Legal experts note that existing privacy laws, such as the California Consumer Privacy Act and Europe's GDPR, may not clearly cover AI-generated derivatives of public images, leaving a regulatory gap that Meta's feature exploited.
Meta defended the feature by stating that users can choose to make their accounts private, and that public accounts should expect broader visibility. However, this argument ignores the transformative nature of AI generation. Unlike a simple screenshot, AI can create entirely new images that incorporate an individual's likeness or style, which feels qualitatively different. The company's quick reversal suggests it recognized this distinction.
Reactions from the Creator Community
Reactions among creators were polarized but largely critical. Some argued that public figures have less expectation of privacy, but many echoed the sentiment that compensation and consent are necessary. 'I spend hours curating my feed to build a certain look. Having an AI rip that off without my permission is just theft,' said a prominent fashion influencer on Instagram. Talent agencies also raised the specter of financial harm: if brands see that any Instagram aesthetic can be replicated by AI, they may devalue the work of original creators.
Conversely, a few artists and AI enthusiasts saw the feature as a powerful creative tool. They pointed out that similar technologies, such as the 'Style Transfer' in other platforms, have been used for legitimate art. However, even they acknowledged that the lack of notification was a major oversight. 'If Meta had implemented a system that allowed public account owners to opt in or out, with proper tagging, it might have been received differently,' one digital artist noted.
The Broader Trend of AI Governance
Meta's misstep is emblematic of the broader challenge facing tech giants as they race to integrate AI across their products. The pressure to stay competitive with rivals like OpenAI, Google, and Microsoft often leads to premature rollouts. In the case of Instagram, Meta was simultaneously rolling out several other AI features, including automated comment replies and AI-generated stickers, which have also generated mixed reactions. The removal of Muse Image may indicate a shift toward more cautious deployment, at least in the short term.
However, skeptics argue that Meta's apology and removal are performative. The company has a history of backtracking after public outrage only to reintroduce similar features later with minor modifications. For instance, in 2024, Meta temporarily disabled a facial recognition feature on Facebook after privacy complaints but later brought it back under a different name. Whether Muse Image will return in a revised form remains unknown, but the underlying technology is part of Meta's long-term AI strategy.
Moreover, the incident highlights the difficulty of regulating AI in real time. Governments around the world are still crafting laws for generative AI, and platforms are effectively operating in a regulatory vacuum. Self-regulation, as demonstrated, is often reactive rather than proactive. The fact that Meta removed the feature only after substantial backlash, rather than during internal testing, suggests that internal ethical reviews may be insufficient.
For now, Instagram users who were concerned about their public photos being used as AI fodder can breathe a sigh of relief. But the fundamental capability remains available to Meta, and the broader ecosystem of AI image generation continues to evolve. Other platforms, such as Snapchat and TikTok, have also experimented with similar features, though with varying degrees of transparency. The incident serves as a reminder that in the age of generative AI, consent and notification are not just nice-to-haves, but essential safeguards that require constant vigilance.
Source:TechCrunch News