mercilessly invade our privacy

Goal has released Muse Image and Muse Videothe first generative AI models for images and videos from its Meta Superintelligence Labs division. With the image model it certainly wants to stand up to GPT Image 2.0 from OpenAI and of course the famous Nano Banana 2 from Google. The most striking thing in fact is not the model itself, which seems competitive in both image and video, but how it has been integrated into Instagram. He did it like a beast and with a clear threat to our privacy… again.

Muse Image and Muse Video come into play. The new models are available for free in the United States from the Meta AI application, and also through Instagram Stories and WhatsApp (where it is available in some additional countries, although they do not detail which ones). It has various effects that can be used directly in Instagram Stories, and on paper performance and quality of both models is promising. The problem is not that, but Meta’s habit of sneaking in its products in a forced way and threatening our privacy.

Just tag yourself. If your Instagram account is public, anyone can write a prompt in Meta AI, tag your username, and generate an image using your real image without you giving your explicit permission for that specific use. Goal presents that option like something casual:

“Whether you want to design a custom event invitation, prototype a collaborative creative concept, or generate a custom graphic, tagging a username allows Meta AI to use public photos to create a post-ready image.”

You will have to opt-out. Meta assumes that you are going to want the entire world to be able to tag you and use your image to create AI images, so if you want to avoid that you will have to be the one to prevent it. In the Instagram help they mention the topicand to prevent your photos from being used like this you have to follow the steps:

1. Enter the Instagram mobile app
2. Tap on your profile and within it on the three lines at the top right
3. Go to “Share and reuse” and look for the “Allow people to create content with yours and reuse it” section
4. There you can disable the existing switches for “Posts” and “Reels”.

Meta doesn’t mention it in the release. In Meta’s official announcements it is indicated that with Muse Image “creative experiences on Instagram and WhatsApp are enhanced”, but the privacy problem is not indicated. Once again Meta simply encourages using these features without thinking that perhaps users have no interest in anyone using their images.

Better to prevent. Instagram makes it clear on its help page that you won’t receive notifications when someone creates AI content using your photos: “you won’t receive notifications about content created using the AI ​​options in Meta,” they explain. Someone could be generating AI images of your face right now and there’s no way to find out unless you stumble across them by chance.

Goal being Goal. Meta’s decision is the same one that has already been seen in Meta—and other companies— on other occasions. We must actively choose to disable certain options (opt-out) instead of giving prior consent (opt-in) for our data to feed the AI ​​models. Google Search for example already stores the content uploads we make, for example when we do a reverse image search, and uses that data to train its AI models.

In Xataka | How to prevent Meta from training its artificial intelligence using your Instagram data