Are dog trackers safe? Data protection in wearables for dogs

This week, half the tech world is debating whether a device on your wrist should be allowed to listen all day long. The same question applies to a device on your dog's neck. Is your dog tracker secretly recording what happens in your home? There is no blanket answer, it depends on the architecture of each product. At Voicari, the answer is no. Voicari is built from the ground up so that only recognized bark events from your dog leave the wearable. Everything else is discarded directly on the device, before anything is even transmitted.

Voicari is an AI-powered smart wearable that recognizes your dog's emotional state, an emotion wearable for dogs. One of the questions we are asked most often is rarely what it can do. It is what it does not do. Below you can read how this architecture works and where you can verify it yourself.

Do dog trackers and wearables constantly listen in on what is said at home?

You are sitting at the kitchen table with your partner in the evening, talking about your day, your dog dozing nearby. It is in exactly these moments that many people wonder whether a device with a microphone might be secretly listening in, whether it is a GPS tracker or an emotion wearable. With many products this is hard to verify from the outside, because it stays unclear what stays on the device and what goes to the cloud.

At Voicari, the answer is no. Because Voicari is always with your dog, we built it to listen only for your dog. The wearable filters out bark events directly on the device. Human speech and ambient sound are discarded immediately: not analyzed, not transmitted, not stored. They never leave the device.

See the data architecture in detail

Data protection through architecture, not through promises

A promise can be broken. An architecture cannot. That is why data protection at Voicari does not begin in a privacy policy, but in the device itself. The filtering runs in a sealed-off area of the chip that neither the app nor Voicari itself can access. The separation therefore happens at the source, before anything leaves your dog's neck.

Voicari creates and stores no audio recordings. What leaves the device is a recognized bark event, not a recording of your home.

• Human speech: discarded directly on the device
• Ambient sound: discarded directly on the device
• Recognized bark events: transmitted to Voicari's servers for emotion analysis

Only then does the actual emotion analysis take place on Voicari's servers. There, the model maps the vocalizations, together with heart rate, respiration and activity, to one of 24 emotional states from 7 core emotional families. This is based on more than 1.16 million data points from over 30,000 dogs across more than 110 breeds.

When your dog barks, that one bark event is recognized and passed on. Your conversation at the kitchen table is not. The separation is that simple, and that is exactly why it can be verified rather than merely promised.

Who processes my dog's data?

Voicari was developed in Berlin and is GDPR-compliant. The emotion analysis runs on Voicari's servers, not with an anonymous third-party provider. Bark events are transmitted encrypted and stored encrypted. The company is part of the Berlin science and startup ecosystem, with links to the Technical University and the Free University of Berlin. For the underlying technology, a patent application was filed with the German Patent and Trademark Office (DPMA) on 27 February 2026.

Who decides when Voicari listens?

You do. You decide whether and when Voicari listens, and you can pause detection anytime in the app. A wearable that is there for your dog should fit around your everyday life, not the other way around.

What does Voicari cost, and when does it arrive?

If you have read this far, you are probably also wondering what all of this costs and when it starts. Voicari Series 1.0 is available as a membership, the device is included: Premium from 276 euros per year, Ultra from 480 euros per year. Founding member spots are limited. The first devices are set to ship from October 2026.

You now know exactly what your dog wears, what the device hears and what it does not, and who processes your data. This transparency is precisely why you can secure your spot as a founding member without having to trust the technology blindly.

You can find all the details about the data foundation, the 24 emotional states and the on-device filtering at voicari.com/research. We also explain how Voicari translates vocalizations into emotional states in How do you know what your dog really feels?. How the device sits on the neck, how much it weighs and how it is charged is shown on the Smart Collar page.

Become a founding member and secure Voicari Series 1.0

Frequently asked questions

Does Voicari record everything that is said in my home?

No. Human speech and ambient sound are discarded directly on the device and never leave the wearable. Voicari creates no audio recordings, only recognized bark events from your dog are transmitted.

Does Voicari store audio recordings?

No. Voicari creates and stores no audio recordings. The device transmits recognized bark events encrypted, and these are stored encrypted.

Can I pause the listening?

Yes. You can pause detection anytime in the app and decide for yourself whether and when Voicari listens.

Is Voicari GDPR-compliant?

Yes. Voicari was developed in Berlin and is GDPR-compliant. The emotion analysis runs on Voicari's servers.

What happens to the bark events that Voicari recognizes?

They are transmitted to Voicari's servers. There, the model maps them, together with heart rate, respiration and activity, to one of 24 emotional states from 7 core emotional families.

Is Voicari a surveillance device for my home?

No. Voicari is built from the ground up for your dog's vocalizations and vital signs, not for human conversations. Those are discarded directly on the device.

Is Voicari a training or anti-bark collar?

No. Voicari is not a shock, anti-bark or training collar and does not correct your dog's behavior. It listens and maps vocalizations and vital signs to an emotional state, nothing more.