What you need to know
- Meta has been pushed to remove Muse Image’s capabilities in its AI chatbot on Instagram following user unrest.
- The AI could snag images of people from public Instagram accounts, turning them into unknown creations based on another user’s prompt.
- Meta grappled with another privacy concern on its smart glasses, as an update now disables the camera if the privacy light is tampered with.
Meta announced the next step for its “people first” AI era last week, but it didn’t sit well with users who saw its intentions for Instagram.
Meta came forward with a statement about its Muse Image AI after extreme user backlash hit the company. Reuters reports that Meta is now “discontinuing” the feature. According to the publication, Meta said, “We’ve heard the feedback that this feature missed the mark, so it’s no longer available,” when referring to Muse Image on Instagram. The company adds that it intended to deliver “a useful creative tool” and to give users “control” over their public imagery.
It’s this last part—the user’s “control”—that’s caused an uproar in its own right. This feature quickly arrived on Instagram last week, but users were opted in for their public content to be used by others, prompting Muse Image in Meta AI. This was done without the knowledge of many, and when they found it, word quickly spread on social media to disable it.
Reuters quotes SAG-AFTRA, the union for actors and more, who said, “Anything other than a clear and conspicuous opt-in for these types of uses of Instagram users’ images is unacceptable…” The union then highlighted the dangers of “nonconsensual digital replicas” as anyone can prompt Muse Image in Meta AI to take someone’s photo and turn them into… anything.
What’s unclear now is how Meta will proceed with Muse Image. The feature is gone from Instagram, but whether we could see it return in a different, perhaps, more acceptable capacity, isn’t clear.
A lack of privacy
There’s one thing to have an AI image generator that “creates” something based on the user’s prompt. But users have been quick to tell Meta that the alternative, using real people’s photos, is a line they’re not willing to cross with it. Muse Image was launched last week and was quickly positioned as an image generator built on Muse Spark—new AI for a new era. One aspect that Meta kept pushing was that Muse Image can be “rooted” in your social circles.
This meant that people chatting with Meta AI on Instagram could @-mention their friend (or anyone with a public profile) and tell it to make something. It’s not necessarily that the person consented to their photos being used. It’s more that, well, their profile is public (or they haven’t opted out), so Muse Image can snag it.
This wasn’t the only privacy concern Meta had to deal with. Last week, the company rolled out a mandatory update that disabled its smart glasses’ camera if the privacy light was tampered with. This update was the result of continued misuse of smart glasses technology, as the public detested “this creepy behavior.”
Android Central’s Take
I’d say I’m surprised. But I’m not. Right off the bat, the feature seemed strange. This is a big feature to opt everyone into. Most people didn’t even realize it was there, or Muse Image’s capabilities. Now that they do, it hasn’t sat well with them. It’s a privacy issue. It’s the feeling you get knowing that someone can take your photo and make you look how they want. It’s unsettling, and now Meta knows it.