Tech Bro Boasts He Used AI to Secretly Monitor Toddlers’ Conversations, Categorize Them In Giant Database

Nicholas Charriere, CEO of AI-powered website builder Mocha, secretly recorded his toddler and her cousin during a sleepover and used Anthropic's AI model Claude to transcribe and categorize their conversations into a private family webpage, boasting about it in an X post on Friday. The post drew widespread backlash, with critics calling the surveillance invasive and disrespectful of the children's privacy, highlighting growing concerns about AI's role in parenting and data privacy.

Parents, if you’re thinking about subjecting your child to some bizarre AI experiment: don’t. “Toddler spent her first night ever sleeping with someone else in the same room: her toddler cousin,” wrote Nicholas Charriere, the CEO of an AI-powered website builder called Mocha, in an X post https://x.com/nichochar/status/2083704833366696118 on Friday. Charriere, seeing that his child was about to have a very formative experience, decided he would record “everything” the two tykes said to each other in their “nonsensical toddler dialect” by “sneaking a mic in there.” Then, of course, he fed the recordings into an AI model, Claude. Using the AI, Charriere explained, he “built a private webpage for the whole family with the audio split into clips, all named and trimmed.” Attached is a photo of the webpage autopsy of the kids’ conversations, divided into a tracklist of all the best goo-goo-ga-ga highlights, like “Hi Mama,” “I’m a Big Boy,” and “I Did Fart.” “Pure joy from everyone,” Charriere declared. “AI is pretty great.” Not everyone agreed. Charriere’s tale was met with heaps of backlash, with many calling it “creepy” and invasive. “I’d hate my parents if I found out they did this,” one replier wrote, receiving more likes than Charrierer’s original tweet as did many other call-out posts . “No respect for privacy. Those are people, not miniature versions of yourself made for the purpose of amusing you.” Mike Duncan, the history podcaster and author, was even more scathing. “Why preserve the sanctity and privacy of late night sleepover chats when you can secretly record everything and feed into the AI slop machine and post about it,” he fumed. Any time parenting and AI intersect is going to be divisive, and it’s a topic we’re going to keep hearing more about. This incident went viral enough that it ended up on X’s trending tab under the incredible AI-generated headline, “Tech Exec Uses AI to Clip Toddler Chats for Family Memories.” Some moms and dads see AI tools as a way to make parenting a little easier, keeping their kids placated and occupied and providing themselves some relief. AI can help them make decisions — like, controversially, picking their child’s name https://futurism.com/artificial-intelligence/parents-named-baby-chatgpt . OpenAI CEO Sam Altman has said https://futurism.com/artificial-intelligence/sam-altman-caring-baby-impossible-without-chatgpt that he heavily relies on ChatGPT to care for his kid and recently mused https://futurism.com/artificial-intelligence/sam-altman-parenting-strategy-horrifying about using AI to generate podcasts on a child’s favorite topics that they can listen to on their way to school. A comedian’s rejoinder, “What if you just talked to your children,” went mega-viral. Other parents say they use ChatGPT’s voice mode to speak to their kid for hours on end https://futurism.com/artificial-intelligence/parents-toddlers-chatgpt-voice-mode . There’s apparently enough demand for this niche that Meta is building an app https://futurism.com/artificial-intelligence/meta-sad-ai-app-stories-children that uses AI to regale children ad hoc, sloppified bedtime stories. AI skeptics, though, see the tech as a lazy way of replacing human interaction and note the privacy risks that AI models pose. AI companies aren’t always transparent about how the data you feed into them, including your conversations, will be used, and it’s possible that the stuff you share will be used to train future AI models. Spying on toddlers and feeding their secrets into an AI might not be giving away anything too compromisingly personal, given how young they are. But you have to admit it’s a little strange to build a website out of it and brag about it online. And how might the child feel when they realize that they can’t trust anything they say to be private? We suspect that there may not be “pure joy” for that revelation.