Meta patents AI device to track emotions and medication intake
The social media giant's expanding patent portfolio raises fresh questions about surveillance, digital identity, and the ethical boundaries of AI in health monitoring.
Meta is reportedly pursuing patent protection for an AI-powered device capable of monitoring users’ emotional states and tracking whether they take their medication. The concept sits at the uncomfortable intersection of health tech and surveillance, two spaces where Meta already has a complicated reputation.
Meta’s long history with emotion detection
This isn’t Meta’s first foray into reading your feelings. The company previously filed multiple emotion-detection patents between 2015 and 2017, exploring techniques like facial expression analysis through cameras, keystroke dynamics, and sensor data to infer users’ moods.
Those earlier patents were designed to enhance content delivery based on emotional state. In English: if the algorithm detected you were sad, it could theoretically serve you content calibrated to that mood.
The bigger picture: Meta’s AI patent ambitions
This filing arrives amid a broader expansion of Meta’s patent portfolio around AI and user interaction. In late December 2025, Meta was granted a patent for a large language model system designed to simulate the social media activity of deceased or inactive users. That patent, initially filed in 2023 and attributed to CTO Andrew Bosworth, essentially creates digital ghosts, AI versions of dead people that continue posting as if nothing happened.
It’s worth noting that other companies are also pushing into the emotion AI space. MetaSoul, for instance, holds patents describing sensor fusion methodologies for detecting emotional states. Feel Therapeutics has explored mental health applications using similar approaches.
What this means for investors and the market
No public connections have emerged between Meta’s patent activities and any token, protocol, or digital asset project.
The regulatory risk is the part investors should watch most carefully. Europe’s AI Act already classifies emotion recognition systems as high-risk in certain contexts. A device that tracks medication adherence and emotional state would almost certainly attract attention from the FDA, FTC, and congressional committees.