The Meta AI app, which has been downloaded 6.5 million times since its launch on April 29, is facing significant scrutiny over privacy issues and the nature of its user-generated content. Social media giant Meta has developed a new app, sparking excitement for how it leverages cutting-edge technology that the company has spent billions of dollars developing. Concerning cases of exposed sensitive personally identifiable information and strange and harmful user-generated prompts have surfaced, casting doubt on the app’s integrity.
TechCrunch, a prominent technology news outlet, obtained screenshots from the Meta AI app that reveal troubling instances of trolling and inappropriate content. Users have been posting private information — such as home addresses and sensitive court information — anywhere on the app. This reckless practice has caused security professionals to sound their warnings. Rachel Tobac, an information security expert, discussed the unregulated sharing of private data between users on the platform.
The app includes a huge variety of user prompts that are incredibly funny and somewhat distressing. One example included a user who generated a picture of Meta CEO Mark Zuckerberg with a pregnant belly. In this fanciful vision, he was performing a wedding ceremony for an insect. One of our users recently asked what to do with red bumps on her inner thigh. This is a wonderful testament to the app’s power to lure in original, creative, out-of-the-box queries.
More examples of confusing user interactions only help to paint a clearer picture of the chaotic environment present within the app. Another asked that we post their phone number in facebook groups to find new dating partners. Another conversation featured a Southern-accented man asking, “Hey, Meta, why do some farts stink more than other farts?” This odd question highlights the app’s capriciousness.
Additionally, essays produced by the app have recently gained notoriety for their portending nonsensicalness. One user prompted the app to generate an image of Mario in a courtroom, including the prompt “super mario divorce.” At the same time, another user with a Pepe the Frog profile picture asked for advice on constructing a water bottle bong. These examples point to the strange and often creepy nature of user interactions on the platform.
Amanda Silberling is a senior writer at TechCrunch, where she writes about the intersection of technology and culture. She’s written a lot about the problems with the Meta AI app. She reflected the widespread concern, among users and experts alike, about the privacy implications of the app’s features.