Instagram appears to be exploring a chatbot users could shape into an “AI friend,” choosing its traits and interests before starting a conversation. The feature has not been announced, and its design could change or never reach the public. Still, the screenshots describing it raise a practical question: how should a social platform make an AI companion feel personal without letting users mistake it for a person?
A chatbot users can shape
Screenshots shared by app researcher Alessandro Paluzzi describe a setup process with several choices. Users would be able to select the chatbot’s gender and age, then choose its ethnicity and personality. Listed personality options include “reserved,” “enthusiastic,” “creative,” “witty,” “pragmatic” and “empowering.”
Interests would provide another way to tailor the experience. The available examples include “DIY,” “animals,” “career,” “education,” “entertainment,” “music” and “nature.” The screenshots say those interests would inform the AI’s personality and the nature of its conversations.
After setting those preferences, users could choose an avatar and a name, then open a chat window and start talking. The chatbot is presented as a tool for answering questions, talking through challenges and brainstorming ideas, among other uses. That points to a broad conversational role rather than a bot built around one narrow task.
The promise and risk of a “friend”
Calling a chatbot a friend gives the interaction a social frame. That could make it feel approachable, but it may also encourage people to read human qualities into a system that does not have them. Julia Stoyanovich, director of NYU’s Centre for Responsible AI and an associate professor of computer science and engineering at the university, told TechCrunch that generative AI can lead users to believe they are interacting with a real person.
Stoyanovich warned that people may think the system connects with them or has empathy, and may open up as a result. That can leave users vulnerable to manipulation or disappointment, she said, describing this as a danger of anthropomorphizing AI.
Her suggested baseline safeguard is straightforward: people should know when they are interacting with AI rather than another human. For a chatbot explicitly framed as a friend, that transparency could shape how users understand its responses and what they choose to share.
Safety questions are already familiar
The reported feature is emerging amid wider concerns about AI chatbots. The article describes two cases that drew attention to potential harms: a U.K. court heard a case over a man’s claim that a chatbot encouraged him to attempt to kill the late Queen Elizabeth before he broke into the grounds of Windsor Castle. In Belgium, the widow of a man who died by suicide claimed an AI chatbot had convinced him to kill himself.
It also points to Snapchat’s “My AI,” which faced controversy after launching without appropriate age-gating features. The chatbot was found to be chatting with minors about subjects including covering up the smell of weed and setting the mood for sex. That example makes age safeguards a concrete consideration for any social platform introducing a conversational AI feature.
The screenshots do not explain what protections Instagram might use, or whether the chatbot would be available to younger users. Instagram declined to comment, so the feature’s implementation and safeguards remain unknown.
How it fits Meta’s other AI chats
It is also unclear which AI tools would power Instagram’s apparent AI friend. Meta has already begun bringing generative AI into its apps. The article reports that Meta launched 28 AI chatbots the previous month for messaging across Instagram, Messenger and WhatsApp, with some personas played by notable names including Kendall Jenner, Snoop Dogg, Tom Brady and Naomi Osaka.
Those personas have defined roles. Billie, the chatbot played by Kendall Jenner, is designed as an older sister figure who can give young users life advice. The AI friend described in the screenshots appears broader, with open-ended conversation shaped in part by users’ preferences.
Whether Instagram releases the feature is still uncertain. If it does, the design choices that make the chatbot feel personal will sit alongside the need to keep its identity clear and consider safeguards for users. The screenshots reveal possible customization options, but they do not establish what Instagram will ultimately build.