In a significant move that blurs the line between social media and artificial intelligence, Meta has quietly launched a new AI image generator called Muse Image. The tool, integrated directly into the Meta AI chatbot, can pull photos from Instagram accounts—including those of other users—to generate custom images. If your Instagram profile is public, anything you've posted could be used as raw material for AI-generated pictures, often without your explicit knowledge or consent.
The release comes as Meta seeks to catch up with competitors like OpenAI and X, both of which launched similar image-generation features in the past two years. But unlike those earlier efforts, Muse Image is deeply tied to the Instagram ecosystem. The chatbot can analyze a user's feed, identify their face, and produce images that mimic their likeness. That means anyone—from friends to strangers—can ask the AI to create a photo of you doing almost anything, limited only by the platform's content policies.
How Muse Image Works
To use Muse Image, you simply log into the Meta AI web app with your Instagram account and type a prompt like “generate a picture of me” or “generate a picture of [another user].” The AI searches through the target account's public photos to learn facial features, hairstyles, and clothing, then outputs a new image. In tests, even private accounts—when logged into by their owner—could be scanned. For public accounts, the process is even more accessible: anyone with an Instagram login can prompt the AI to create an image of another user, whether or not they follow that person.
When journalist Mike Pearl tested the feature for a public Instagram account of a friend he doesn't follow, Muse Image successfully generated an image without the friend's permission. The tool even produced images of Meta CEO Mark Zuckerberg, using only publicly available photos from his Instagram feed. While the generated images in these tests were harmless—generic portraits or neutral scenes—the implications are far-reaching.
Privacy Settings and Default Opt-In
According to Meta's updated help page, public account holders have automatically opted into this feature. The company's policy now states that “people may be able to create content with your Instagram content using AI features at Meta.” This includes not only images but also reels, posts, and stories. The setting is enabled by default for all public accounts, meaning users must actively opt out if they don't want their photos used as AI training material.
The only way to completely prevent your public photos from being accessed by Muse Image is to either change your account to private or manually disable the “Allow people to use your content” toggle in the Instagram app's settings. On desktop, the quickest route is making the account private. On mobile, users can navigate to their profile, tap the hamburger menu, and find the “Sharing and reuse” section. There, an option labeled “Allow people to use your content on Instagram and with AI features on Meta” can be turned off for both posts and reels.
Meta has stressed that Muse Image includes built-in safeguards to prevent the generation of violent, sexual, or defamatory imagery. A company spokesperson stated that content violating community standards will be removed if reported. However, critics point out that enforcement is reactive, and the system's reliance on public data raises questions about consent and digital identity. The feature is currently live for U.S. users on Instagram Stories, with plans to expand to WhatsApp and Facebook.
Historical Context and Industry Parallels
This is not the first time a social media giant has faced backlash over AI-generated imagery. In late 2024, X (formerly Twitter) integrated Grok's image generation features, which were quickly abused to create sexualized images of public figures and even minors. The controversy forced X to implement stricter content filters, but the damage to public trust had already been done. OpenAI's Sora video generator, launched in September 2025, similarly faced privacy concerns and was shut down by March 2026 as the company pivoted focus.
Meta's late entry into this space might be an attempt to learn from those mistakes, but early tests suggest the same vulnerabilities persist. The ability to generate images of strangers without their permission echoes the deepfake problem that has plagued the internet for years. While Meta's safeguards may block the most egregious abuse, they do not address the core issue: using someone's likeness as a creative resource without their active consent.
In the broader tech landscape, the line between user-generated content and AI training data continues to erode. Earlier this year, LinkedIn was found to host a significant portion of AI-generated posts, blurring authenticity on the platform. Meta's foray into AI image generation from Instagram feeds may mark a new chapter in that trend, where every public photo becomes potential fuel for a machine's imagination.
How to Protect Yourself
For those uncomfortable with their images being used by Meta's AI, the steps are simple but require immediate action. Besides making your account private or toggling the sharing settings, users can also delete old public posts or watermark their images. However, once an image is public and has been scraped by the AI, there may be no way to fully retract it. Privacy advocates recommend regularly reviewing your account's visibility settings and staying informed about platform changes.
The rollout of Muse Video, expected later this year, will only amplify these concerns. While the still-image generator can already create convincing portraits, a video generator would enable even more invasive uses. Meta has not disclosed how it plans to handle consent for video content, but the default opt-in approach suggests the company is prioritizing feature adoption over user autonomy.
As AI tools become more integrated into everyday social media, the debate over digital rights will only intensify. For now, the onus is on users to protect their own likeness—a task that becomes increasingly difficult as the technology evolves.
Source:Gizmodo News