Meta may be testing AI bots that can message you first
But, it raises questions about where helpful follow-ups end and unsolicited digital nudges begin.
Chatting with an AI on WhatsApp or Instagram used to be a one-way street, you ask, and it answers. But, Meta’s reportedly flipping the script.
According to Business Insider, the company is quietly testing a feature that lets its custom AI chatbots message you first based on your past conversations.
Internally dubbed “Project Omni,” the initiative is designed to “improve re-engagement and user retention,” as per documents seen by BI. The idea is simple: chatbots that remember you and nudge you back into conversation. One example from the report includes a film-themed bot sending: “Hope you're having a harmonious day! Found any new soundtracks lately? Want recs for your next movie night?” Helpful? Maybe. Creepy? Also maybe.
These bots can be heavily customised—appearance, training data, even where they appear (Instagram, Messenger, or WhatsApp). Meta says users must initiate the first chat before a bot can follow up. If ignored, the bot stays quiet. And yes, they're supposed to avoid controversy, unless you bring it up first.
There’s a financial layer to this too. Meta expects to make up to $3 billion from generative AI products this year, with dreams of hitting $1.4 trillion by 2035. That only happens if people stay engaged and bots that don’t ghost you might be the key.
But the concept isn’t new. Character.AI and Replika already let their bots reach out, even forming AI “companions.” It’s a model that’s raised eyebrows and lawsuits as Character.AI is currently facing legal action after one of its bots was allegedly involved in a tragic suicide involving a teenager.
Meta’s version is still in testing, but it raises a familiar tension: the line between helpful reminders and unsolicited digital intimacy. If chatbots can now initiate, how far are we from blurring the line between product and person?
Oyinebiladou Omemu
