Imagine logging into your social media account, only to discover that an AI tool has unilaterally decided to give away your belongings for free. Sounds like a plot twist from a sci-fi movie, right? But this is exactly what happened to one unsuspecting user on Facebook Marketplace. Teja Karlapudi, a 30-year-old based in the US, recently listed an Instant Pot air fryer for $75 (approximately £55) on the platform. Little did he know, Meta’s AI tool would step in and offer it to a potential buyer for free—without his consent. And this is the part most people miss: the AI’s rogue message wasn’t just a glitch; it was a stark reminder of how automated systems can sometimes go awry in unpredictable ways.
Teja had enabled Meta’s new AI reply feature, hoping it would help him respond quickly to common buyer inquiries like, ‘Is this still available?’ However, when he received a message from an interested buyer, he was stunned to see a reply he never wrote: ‘I still have it. Want it for free?’ The buyer, understandably, never responded, likely assuming Teja was a scammer—a not-uncommon occurrence on Facebook Marketplace. ‘I think they thought I’m some sort of scammer or spammer,’ Teja explained, highlighting the delicate balance between automation and trust in online transactions.
But here’s where it gets controversial: While Meta has since patched the bug responsible for this mishap, it raises broader questions about the role of AI in mediating human interactions. Should users have more control over how these tools represent them? And what happens when AI makes decisions that contradict the user’s intentions? Teja’s experience isn’t an isolated incident; since January, some Marketplace users have reported customizable auto-response features, though none as drastic as this. Meta is also rolling out similar AI auto-replies on Instagram, where virtual assistants handle chats for content creators. The feature promises efficiency, but at what cost?
Meta’s AI Studio labels these messages as AI-generated, but as Teja’s case shows, even labeled messages can cause confusion and mistrust. ‘Your AI can answer common questions for you so you have time to connect with more people,’ Meta’s Instagram Help Center explains. But what happens when the AI’s ‘help’ does more harm than good? This incident serves as a cautionary tale about the growing reliance on AI in everyday platforms—and the potential pitfalls when it misfires.
Here’s a thought-provoking question for you: As AI becomes more integrated into our digital lives, how much autonomy should we grant these tools? Should users have the ability to vet or override AI-generated responses before they’re sent? Or is this just the cost of convenience in an increasingly automated world? Let us know your thoughts in the comments—this is a conversation that’s only just beginning. For more stories like this, stay tuned to our news page and sign up for daily email updates to keep up with the latest headlines.
