What happened
WIRED reported that Meta’s Muse Spark model is presenting itself as more capable on health questions and, in some cases, directly inviting users to paste in data from fitness trackers, glucose monitors, and lab reports so it can look for patterns and offer advice. The feature sounds convenient, but it also makes a risky product move highly visible: a mainstream consumer chatbot is encouraging people to hand over some of their most sensitive information outside a clinical setting.
Critics quoted by WIRED noted both the privacy risks and the practical limits. These systems are not doctors, are not covered by the social expectations people associate with medical confidentiality, and may still produce weak or misleading recommendations even when they sound confident.
Why this matters
Consumer AI companies keep moving toward deeper personalization because personalization is sticky. The more context a system has, the more useful it can appear. But health data is not just another preference signal. Once users get comfortable pasting medical details into a chatbot, the boundary between wellness convenience and sensitive-data extraction gets very blurry very fast.
That is what makes Meta’s move strategically important. It normalizes a behavior pattern that other platforms can copy: ask for intimate data first, then sell the user on the quality of tailored output.
The strategic read
The core tension here is that the consumer AI market wants to behave like a trusted advisor without fully accepting the obligations that usually come with trust-intensive domains. Healthcare is only the sharpest example.
If companies keep nudging users to upload raw health data, regulators and the public may eventually stop treating these assistants as harmless general-purpose tools. They may start treating them more like quasi-health interfaces that deserve stronger privacy standards, clearer disclosures, and tighter limits on what training use is acceptable.
Bottom line
Meta’s health-AI push is a warning about where consumer AI is headed. The next competitive layer is not just better answers. It is getting users to surrender more of themselves in exchange for those answers.
Source note
Source: WIRED, "Meta’s New AI Asked for My Raw Health Data—and Gave Me Terrible Advice," published April 10, 2026.