Meta Files Patent To Track Your Conversations All Day Under The Guise of Emotional Wellbeing

A newly published Meta patent describes AI that could analyze your voice, estimate your emotional state and connect it to where you are, what you're doing and more.
We receive compensation from the products and services mentioned in this story, but the opinions are the author's own. Compensation may impact where offers appear. We have not included all available products or offers. Learn more about how we make money and our editorial policies.

Your smartwatch counts your steps. Your phone tracks where you go. Now, imagine an AI that keeps tabs on how you feel.

That's the idea behind a newly published Meta patent application describing a system that could analyze a user's voice over time to estimate their emotional state. The filing describes wearable devices capable of the "continuous or intermittent capture of audio, video, and motion data" to help AI understand how users feel over time, which could lead to persistent data collection without explicit consent.[1]

Meta says the system could identify patterns such as someone being happier after taking medication, sighing more frequently between meetings, or feeling calmer at certain times of day by correlating voice recordings with contextual information like location, activities and digital interactions.

In this article
Meta envisions AI that listens for more than just words
What Meta's after
Could Meta actually collect this information?
Why you should pay attention
How to protect your privacy

Meta envisions AI that listens for more than just words

Rather than focusing only on the words a person speaks, Meta says an AI assistant could also pay attention to nonverbal cues.

One example in the patent describes a smart home device that records a user's sighs, timestamps them and analyzes them alongside speech to determine emotional trends over time. Another describes smart glasses that capture spoken conversations and present emotional summaries based on what the wearer says and how they say it.

The proposed system would also transcribe your conversations, analyze them using an emotional-state machine learning model and correlate those findings with contextual information, including:

  • Time of day
  • Physical location
  • Current activity
  • The digital service or app being used

The result would be summaries showing emotional trends over time, complete with "citations" referencing portions of a user's own conversations that support the AI's conclusions.

What Meta’s after

Unlike mood-tracking apps that rely on users manually logging how they feel, Meta argues current tools leave "significant gaps" because they depend on self-reporting and "short, isolated interactions."

Instead, the patent proposes collecting information over longer periods to create a more complete picture of emotional wellbeing. The filing says the AI assistant could "track emotion over a predetermined time period (e.g., a month)" before generating summaries of a user's emotional state.

The patent even provides examples of the kinds of patterns the system could uncover, such as identifying "a happier emotional state associated with a particular time of day or at a time when medication is taken," or spotting increased sighing between meetings.

Could Meta actually collect this information?

For technology like this to work, it would require access to a user's microphone and potentially other permissions, including location data. Being aware of and managing these permissions is crucial for maintaining control over your privacy, as apps on both Android and iPhone must receive permission before accessing a device's microphone, which users can revoke through privacy settings.

Meta already collects audio that users intentionally provide through voice-enabled features, according to the company's privacy policy. That differs from the system described in this patent, however.

While the filing discusses devices that capture audio during everyday use, Meta has not announced a consumer product that continuously monitors users' conversations to build long-term emotional profiles. A patent application simply protects an invention and doesn't necessarily indicate the company plans to release it.

Why you should pay attention

The patent repeatedly references fitness coaching and emotional wellness as potential uses for the technology.

But it also describes an ecosystem of AI-powered devices that analyze speech, laughter, sighs, tone of voice and other behavioral signals throughout the day, then combine them with contextual information to produce a long-term emotional dossier about you. 

The information Meta could collect with this patent would be a gold mind of personal data they could use to sell to companies or for their own internal advertising. This brings that creepy feeling that your phone is listening to you to a whole other level.

How to protect your privacy

Even though the technology described in Meta's patent isn't available today, reviewing your privacy settings is still a good habit.

You can reduce unnecessary data collection by:

  • Reviewing which apps have microphone permission.
  • Removing microphone access for apps that don't need it.
  • Limiting location permissions where possible.
  • Periodically checking app permissions to ensure they still match how you use your device.

Take Control of Your Online Privacy
5.0
Editorial Rating
Get Deal
On Incogni's website
2026 Editors’ Choice
Best Overall Data Removal Service
Privacy Protection
Incogni
PROMOTION: Save 55% with code COOKIE
  • Top-rated data removal service that scrubs your info from 420+ data broker sites automatically
  • Independently verified by Deloitte, meaning removals are actually sent, confirmed, and not just claimed
  • The Unlimited plan extends coverage to 2,000+ additional sites with human-assisted custom removals

Author Details
Thomas Kent is a multi-disciplined reporter with over a decade of experience covering online platforms, digital trends, and consumer-facing tech. Tom focuses on digital privacy, data tracking, and user behavior, with a particular interest in how cookies, online surveillance, and platform design shape the modern internet experience. His reporting takes a research-driven, news-focused approach, translating complex technical topics into clear, accessible insights.

Citations

[1] EMOTIONAL STATE ANALYSIS AND REAL-TIME FITNESS COACHING