A new patent filing by **Meta Platforms** has unveiled plans for an AI system designed to continuously monitor and log users' emotional states. The application, **US 2026/0182881**, filed in December 2025 and published on July 2, outlines a technology that listens to your voice, analyzes your tone, and records your perceived feelings, timestamping each entry with contextual data like location and device activity. While **Meta** has not announced any product incorporating this technology, patent filings often stake a claim on ideas long before development begins. The patent names **Lachlan Dunn** as the sole inventor and was first flagged by the patent-analysis site **Patentlyze**. ### What the Patent Describes The proposed system would use devices such as smart glasses, phones, smartwatches, headphones, or smart home speakers to record speech throughout the day. An AI, trained to interpret mood from both spoken words and vocal characteristics (tone, pace, sighs, laughs), would then tag audio segments with emotional reads. These emotional reads would be correlated with contextual information, including the user's location, current activity, and even how they are interacting with their phone. Over time, the system would compile a summary of emotional patterns, offering insights like, "You sigh most frequently before bed, and you're happiest when with friends. You've expressed more gratitude this month." Crucially, the filing indicates the system's reach extends beyond voice. It can integrate biometric and eye-tracking signals, such as pupil size, blink rate, and eye moisture, to detect stress or crying. It can also monitor device usage, including viewed posts, screen time, and app switching speed, all feeding into a unified emotional profile.  ### Integrated Fitness Coaching The patent also describes a fitness coaching component. Smart glasses could monitor a user's form during exercise, providing real-time feedback and encouragement. This coach would adapt to the user's detected mood, easing off if discouragement is sensed, or even "admonishing" the user if they are perceived to be slacking with ample energy.  ### Past Precedents and Regulatory Landscape This isn't the first foray into voice-based emotion detection. **Amazon** previously included a 'Tone' feature in its **Halo** wearable in 2020, which analyzed pitch and pace to provide mood feedback. However, **Amazon** processed these samples on-device and deleted them without cloud storage. Despite this, it faced intense scrutiny, with Senator Amy Klobuchar raising privacy concerns over the collection of such personal health data. **Amazon** ultimately discontinued the **Halo** line in 2023. **Meta's** patent expands significantly on **Amazon's** approach by incorporating eye-tracking and device usage data, creating a far more intrusive profile. While some versions of the patent suggest on-device processing, others explicitly mention logging data to servers. Regulators are already addressing the complexities of AI-driven emotion detection. The **EU's AI Act**, effective February 2025, bans AI that infers emotions in workplaces and schools, citing the unreliability and cultural variability of emotional expression. Fines for non-compliance can reach 35 million euros or 7% of global turnover. While this ban currently excludes consumer tools, a separate rule in August 2026 will mandate disclosure for systems inferring emotions from biometric signals. A voice-first coach, especially one integrating pupil and blink rate analysis, would likely fall under this biometric disclosure requirement. **Ghost Protocol** has reached out to **Meta** for comment on potential product plans and data handling practices, and will update this story with any response.