Anyone who rode the New York City Subway in the last few months of 2025 knows how this idea worked out. Friend, which ran a massive marketing campaign throughout the city’s subway system to promote its AI pendants, received widespread public backlash over its billboards, which many interpreted as promoting artificial relationships rather than real, human-to-human connections. Advertisements were widely defaced throughout the city. (A memorable piece of graffiti advertising a friend saw in Brooklyn in October read simply: “Call your mom.”)
Whether the Friend’s Subway marketing debacle was a net loss for the company is debatable; As the old saying goes, all publicity is good publicity. But what doesn’t seem debatable at this point is that overall most people weren’t thrilled with the idea of wearing and monitoring a portable AI chatbot. Also witness the fall of Humane’s AI pin, which was similarly hyped for a brief period as a harbinger of a major new tech trend, before it too failed to catch on with consumers. Heumann later sold its assets to HP.
The dream of wearable AI is not dead
However, given the failure of the AI Pin and the public reaction to the Friend, it would be premature to conclude that AI wearables are a lost cause. After all, it’s been less than four years since the public debut of ChatGPT, the starting gun that started the AI race. Only a very small handful of tech developers have managed to monetize their proprietary LLMs, meaning there are still a lot of unanswered questions about what the next generation of AI-powered products will look like.
Silicon Valley also includes major players, such as some Apple And Meta continues to believe in the practical and commercial potential of AI-powered wearable devices. But whatever form those new wearable devices take, they almost certainly won’t be “companions.” come here Friendly and humanitarian. The public’s verdict on that model has been absolutely clear.
An alternative path that the development of AI wearables could take is toward personal health. In fact, you can see that development is already happening.
ora ring 5Available for preorder today and starting at $399, comes with a built-in AI assistant our advisor (Which the company first launched last year). While most people associate the Ora ring with sleep tracking, the new smart rings are largely being marketed as personal fitness trackers and health coaches. Through integration with the Oura app, users can get personalized health recommendations from an AI assistant as well as human medical professionals.
The launch of the Ora Ring 5 follows Google’s release of the Fitbit Air, which comes with a personal, LLM-powered assistant called Health Coach. (Google acquired Fitbit for a reported $2.1 billion in 2021.) It also comes with a new, AI-heavy app called Google Health, which Couldn’t get the best start.
Other AI-powered wearable devices geared toward health optimization abound. For example, some smart headbands perform EEG scans while you sleep to measure sleep quality. And Meta’s Ray-Ban smart glasses — perhaps the most popular (and controversial) AI-focused wearable device on the market at the moment — come with a few health-oriented features, including one that allows you to Keep track of everything you eat.
chatbot health assistant
Oura and Fitbit, of course, have always been geared towards health and fitness, and it was only a matter of time that they started leaning more into AI, so we shouldn’t be surprised if built-in AI health assistants are becoming a new trend in the wearable tech sector. However, at the same time, some leading AI developers have begun to invest more in personal health features, in direct response to trends in users’ interactions with chatbots.
For example, in December, a report Published by Microsoft revealed that “health and fitness” was the third most common category of user prompts fed to CoPilot, the company’s flagship AI chatbot (the first and second most common types of questions were “technology” and “work and careers”, respectively); Microsoft introduced copilot health in March.
OpenAI data published in January axios showed Forty million people worldwide were using ChatGPT for medical advice; OpenAI launched chatgpt health And ChatGPT for Healthcare Later that week.
risks and opportunities
There are a lot of legitimate concerns about using AI as a personal health coach, whether it’s through a wearable device or a chatbot on your phone. Chief among them: What’s actually happening with that data? How is it being used by the companies that own the devices, who can monitor it and to what extent? And in the case of chatbots, the risk of hallucination – giving users fake information as if it were fact – increases when the conversation focuses on personal health.
Like other hot-button legal and ethical puzzles posed by AI, such questions will take time to resolve. But the growing interest in personalized health among tech companies indicates that LLM has the potential to become a key area of investment for both leading AI labs looking to monetize and new startups with an eye on the next big tech trend.
In the meantime, all those companies will be guided by the lessons of Friend and Humane’s AI pin: If AI wearable devices ever go mainstream, it will almost certainly not be in the form of a “companion”, but as a more limited device, designed for a more specific purpose. For better or worse, many people will still turn to AI for therapy, relationship advice, and yes, even as a balm for loneliness. But most of them do not want to have their personal issues publicly advertised in some specific gadget hanging around their neck.
<a href