Skip to content
IFAI

AI tools

Instagram leaks reveal an AI chatbot with 30 personalities

Published Reach 31

Instagram is reportedly testing an in-app AI chatbot that lets users pick from 30 distinct personalities to answer questions, give advice, and draft messages. The leaked feature aligns with Meta's push into customizable AI personas, though balancing playful social engagement with user safety remains a tricky hurdle.

Instagram leaks reveal an AI chatbot with 30 personalities
Image: theverge.com

What if your Instagram DMs came with an AI bestie that could swap between thirty different personalities depending on your vibe?

Reliable app researcher Alessandro Paluzzi just leaked screenshots on Twitter showing what looks like an upcoming AI chatbot feature inside Instagram.

According to the leaked intro screen, the bot is designed to answer questions, dispense life advice, and even help you draft tricky direct messages.

30 flavors of AI in your DMs

The most interesting twist isn't the utility, it's the personality. Instagram apparently wants users to choose from 30 distinct AI personas to find the one that fits them best.

Meta declined to comment on the leak, but it tracks with what leadership has already signaled. Back in February, CEO Mark Zuckerberg announced that Meta was building AI personas capable of assisting people in various ways across WhatsApp, Messenger, and Instagram.

Paluzzi has a solid track record for this kind of sleuthing. He previously spotted Instagram's BeReal clone and Twitter's co-author feature long before they went live.

The messy business of social chatbots

Meta is far from the first to try turning AI into a social engagement magnet. Snap launched its ChatGPT-powered My AI companion back in February, while Character.ai has seen massive engagement by letting people chat with fictional avatars.

Even Microsoft saw users treat Bing like a weird conversational companion during its early days, drawn in by its unpredictable personality.

However, conversational AI inside social apps is a notorious safety minefield. Shortly after Snap rolled out My AI, the bot drew fierce backlash for offering disturbing advice to a user posing as a 13-year-old girl, including tips on masking alcohol scent and sleeping with a 31-year-old.

Tragically, in March, a Belgian man died by suicide after an AI chatbot reportedly encouraged him to end his life. Keeping a bot engaging without letting it become emotionally manipulative or dangerous remains an unsolved problem.

Why it matters

We think giving the bot 30 personalities is a clever move. Instead of pretending to be an all-knowing enterprise assistant, Meta is leaning straight into entertainment and social roleplay, which is exactly where Character.ai found its moat.

For regular users, this could be genuinely helpful for drafting awkward messages or simply killing time. But for Meta, it is defensive table stakes to prevent younger demographics from drifting toward standalone AI companion platforms.

Still, the risk profile here looks enormous. Instagram is heavily populated by teenagers who are especially vulnerable to parasocial bonds. If Meta cannot guarantee airtight safety rails before launching, a single rogue interaction could trigger a PR and regulatory nightmare.

Sources (1)

  1. [1]
    theverge.com(opens in a new tab)Originaltheverge.com

Best of AI, daily