Instagram’s New AI Image Generator: What It Does and How to Opt Out
In early July 2026, Instagram quietly launched a feature powered by AI that lets anyone take a public photo and have the tool generate a new image from it—altering backgrounds, changing outfits, or even placing the person in a scene they were never in. For many users, the fact that this tool works on their pictures without their explicit permission came as an unwelcome surprise. Here’s a breakdown of what the feature does, why privacy experts are concerned, and exactly how to disable it for your account.
What happened
The new capability, which Meta (Instagram’s parent company) calls “AI image generation,” allows a user to upload or select any photo from a public Instagram account and then apply text prompts to transform it. The resulting image is a synthetic edit that looks realistic, and it can be shared within the app.
According to reports from The Guardian and Yahoo Finance UK, the feature is enabled by default for accounts that are set to public. If your profile is public, your images are automatically available for the AI to work with unless you opt out. Only private accounts are excluded by default.
Meta’s existing terms of service already grant the company a broad license to use user content for “product improvement,” but this new generator represents a direct, interactive use of personal images by other people—not just by Meta itself.
Why it matters
Privacy and security experts have flagged several risks. The most immediate is the ease with which a deepfake can be created from a real photo. Someone could take a picture of you from your feed and generate an image that makes it look like you’re doing something you never did, in a place you’ve never been. This can be used for harassment, reputation damage, or blackmail.
In a related article, The Guardian highlighted that UK schools are being advised to remove pupils’ online photos because the threat of AI-generated blackmail is growing. Meta’s tool lowers the barrier for such abuse even further.
There is also the question of consent. Because the feature is opt-out rather than opt-in, many users are unaware their content is being used until after the fact. Even if you delete a post later, the AI may have already been trained on or derived from your image.
What readers can do
The opt-out exists, but it’s not immediately obvious. Here’s how to turn it off on Instagram:
- Open the Instagram app and go to your profile (bottom-right icon).
- Tap the hamburger menu (three horizontal lines) in the top-right corner.
- Select Settings and privacy.
- Scroll down to Privacy and tap it.
- Look for a section labeled AI (or search “AI” in the settings search bar if you don’t see it).
- Toggle off “Allow others to use your photos with AI image generation” (the exact wording may vary by region and app version).
Repeat these steps for any connected accounts (e.g., Threads, if you use it). The change should take effect immediately, but it won’t retroactively remove images that were already used.
What if your images have already been used?
There is no built-in way to delete AI-generated images that others have already created from your photos. You can report specific generated images if they are abusive (tap the three dots on the post, select “Report”). However, this only addresses individual instances, not the underlying availability of your content. For stronger protection, consider switching your Instagram account to private. Private accounts are not affected by the feature, according to the same reports.
Broader implications
This feature is part of a wider trend where social media platforms are integrating generative AI directly into user interactions, often with weak default privacy settings. The regulatory landscape is still catching up; the EU’s AI Act and other frameworks may eventually require opt-in consent for such tools, but right now, the responsibility falls largely on the user.
If you post publicly on Instagram, assume your images are available for AI manipulation unless you actively turn it off. Share these steps with friends and family who may not be aware—especially those with public accounts who post photos of themselves or their children.
Sources: The Guardian (July 9, 2026), Yahoo Finance UK (July 10, 2026), inkl (July 9, 2026).