For users to be comfortable with AI, they need to know when they are interacting with it. This is not the case in two recent incidents.
The first involved a covert experiment on Reddit conducted by researchers at the University of Zurich, who deployed 13 AI-powered bot accounts. These bots posed as people with sensitive backgrounds—like trauma counselors and assault survivors—and crafted personalized replies based on users’ posting histories. The goal was to study whether AI or humans were more persuasive. Despite being approved by the university’s ethics board, the backlash from the Reddit community and academic institutions was swift. Reddit’s Chief Legal Officer condemned the study in an official post on the affected subreddit.
In a similar—but less serious—case, a rhythmic CHR radio station near Sydney, Australia was caught passing off an AI-generated midday DJ named “Thy” as a real person. After six months on-air, a Substack gumshoe hinted about the host’s authenticity, leading to a wider exposé in The Australian Financial Review.
In both cases, the issue isn’t the use of AI itself—there’s real value in exploring AI’s capabilities and even embracing AI personalities. The problem is deceiving users, intentionally or not, about who—or what—is on the other end of the interaction. That kind of secrecy crosses an important line.
Meta Releases New AI App During LlamaCom
The Facebook/Insta/WhatsApp parent held its first developer conference dedicated to news and releases around Llama, its AI platform. Highlights were sparse but included the release of a new stand-alone AI app with functionalities similar to ChatGPT and Gemini and Copilot. Other announced changes were cringy adjustments to the privacy policy for Meta smart glasses users that reduced control over their data. [details]
Google Inserts Ads Into AI Chatbot Conversations
As AI increasingly answers user questions directly, reducing traditional website traffic, Google has started placing ads for third-party products within AI chats. Google told Bloomberg, “AdSense for Search is available for websites that want to show relevant ads in their conversational AI experiences.” [details]
Why Call Mom When AI Can Do It for You?
Nothing says “I love you, Mom” like outsourcing affection to a robot.
In a plot twist ripped from Black Mirror, InTouch has launched a “Call Mom” subscription service that promises “meaningful, mind-stimulating conversations with your parent”—so you don’t have to bother.
Journalist Joseph Cox of 404 Media had his own mom try it out. Her review? “It sounds from the clips like talking to an automated company cold call.” Not exactly the Hallmark moment shown in the company’s homepage image—a glowing father-daughter scene featuring a suspiciously mangled left hand, courtesy of AI art.
In a world where deepfake scams already prey on the elderly, teaching Grandma to chat with chatbots feels less like innovation and more like a terrible idea wrapped in a subscription fee.
Originally published by Jacobs Media
