AI is becoming more personal, not just helping write emails or writing code or summarize meetings anymore.
Products are starting to understand conversations, routines, memories, and everyday experiences in ways that would have felt impossible only a few years ago. But the closer technology gets to daily life, the more important privacy becomes. Naturally, that changes the kinds of questions people ask.
- What exactly is being captured?
- Who owns the data?
- Can I delete everything?
- Is this training AI models?
- How private is this, really?
For Looki, these questions were not something we wanted to address after building the product. Privacy is a big part of shaping how the product was created from the beginning.
The company was founded by two fathers who understood firsthand what it means to build a device that centers on personal life, family time, and holidays/vacations. From day one, there was a strong awareness that a product designed around memory would inevitably live close to the parts of life people care about most: family conversations, ordinary routines, moments with children, quiet experiences that may not seem important at the time but become meaningful later.
Those moments do not feel like “content.” They feel personal.
That perspective shaped many of the decisions underneath Looki itself.
Built To Feel More Natural In Everyday Life
One of the first privacy decisions we made was around form factor.
Many AI wearables today are built around glasses. While convenient in some situations, glasses can also feel socially ambiguous. People around you often cannot tell when they are active, recording, or paying attention.
Looki L1 was intentionally designed differently.
Because it is not attached to your face, users can naturally remove it, clip it inward, place it down, or turn it away depending on the environment. In practice, this makes the device feel more flexible and socially comfortable in everyday life.
We also designed the product with visible trust cues and clear boundaries in mind. Looki uses visual indicators and system sounds to help communicate device activity more clearly, so interactions do not feel hidden or ambiguous.
When the camera is intentionally covered or facing inward, visual AI functions automatically stop working. The idea is simple: privacy controls should feel physical and intuitive, not buried inside menus or settings.
We think wearable AI should adapt to human environments — not force people to adapt to the technology.
Content Stays Local By Default
By default, captured content is stored locally on the device.
You choose whether to sync content or turn on a feature that needs cloud processing. Some features may automatically upload the content they need while they are turned on.
We felt strongly about this from the beginning. A product designed around personal memory should not assume everything belongs in the cloud automatically.
When content is successfully uploaded, the copy stored locally on the Looki L1 is scheduled for automatic deletion 12 hours after the upload is completed. This device-side cleanup should not be confused with deletion of content stored in the user’s Looki account. Content the user chooses to save or synchronize, together with Moments, Memories, Journal entries, AI conversations, Vlogs, comics, summaries, and other generated content, may remain available until the user deletes it or closes the account.
Users always remain in control of what gets processed, kept, and deleted. For us, local-first storage was an important principle from the start.
Sensitive Content Is Filtered Automatically
Looki L1 uses AI filtering systems designed to identify sensitive environments and prevent certain types of content from being uploaded for cloud analysis. That includes spaces where users would reasonably expect stronger privacy protections, such as bathrooms or other highly personal environments.
As AI becomes more integrated into everyday life, products need to behave more responsibly around real human boundaries.
Personal Content Is Not Used to Train AI Models
One of the clearest boundaries we believe personal AI products should have is around training data.
Looki does not use raw or identifiable personal content to train third-party general-purpose AI models. Looki also does not currently operate its own foundation or fine-tuned models. Its current AI features are provided through third-party API services subject to contractual purpose limitations, zero-training requirements, and applicable zero-data-retention configurations.
Your family moments, conversations, routines, and experiences are not treated as generalized training material for systems serving everyone else.
As AI becomes more integrated into everyday life, this distinction becomes increasingly important.
Memory is personal. AI should treat it that way.
Encryption in Transit, at Rest, and Controlled Cloud Processing
Your data is encrypted when it is sent and while it is stored in Looki’s cloud. Access is strictly limited. When you use a cloud-based AI feature, Looki and its authorized AI providers may temporarily decrypt the content needed to make that feature work. Your content is used only to provide the feature you requested.
The system is also designed to comply with applicable local privacy and data regulations, depending on the region.
Most consumers may never think deeply about infrastructure architecture, but they do care about the underlying principle: personal experiences should be handled carefully and securely.
We agree.
Users Stay In Control
Ultimately, privacy is not only about encryption or infrastructure.
It is about control.
Users can review, manage, and delete eligible captured content through the Looki App. Personal experiences should never feel trapped inside a system once they are recorded.
As AI becomes more contextual and memory-based, we think ownership needs to mean something concrete:
-
visibility
-
control
-
clear boundaries
-
understandable behavior
Not vague promises hidden behind technical language.
We do not think privacy should feel hidden behind long policies or buried inside settings menus.
Users should understand how the product behaves. They should know what gets stored locally, what gets uploaded, and what stays under their control.
As AI becomes more integrated into daily life, products will increasingly be judged not only by what they can do, but by how responsibly they behave around the people using them.
For us, that starts with a simple principle: Personal memories should remain personal.
To learn more, you can read our privacy policy here.
Frequently Asked Questions
Is Looki always recording?
You are always in the driver’s seat. Looki offers multiple modes designed around different levels of capture and interaction, so users can choose what feels appropriate for their environment and comfort level.
In AI Mode, Looki uses contextual sensing to better understand and respond to what is happening around the user in real time, enabling experiences like contextual recall, memory assistance, and AI interactions.
At the same time, Looki was designed with clear boundaries and user awareness in mind. The device uses visible trust cues, system sounds, local-first storage, and intuitive physical controls to help make its behavior more understandable and transparent. When the camera is intentionally covered or facing inward, visual AI functions automatically stop working.
You can also review, manage, and delete eligible captured content through the Looki App. If Looki captures something you would rather keep private, you can remove it using the available content controls.
Is Looki cloud-based?
Content is stored on the Looki L1 first. You can manually upload content or turn on a feature that needs cloud processing. Some features may automatically upload the content they need while the feature is active.
Does Looki use personal data to train AI models?
No. Looki does not use your personal photos, videos, audio, or other identifiable content to train third-party general-purpose AI models.
Can users delete their content?
Yes. You can review, manage, and delete eligible captured content through the Looki App. Content saved or synced to your account may remain there until you delete it or close your account.
How does Looki protect sensitive content?
Looki uses AI filtering systems designed to identify sensitive environments and prevent certain content from being uploaded for cloud processing.
Is the uploaded data encrypted?
Yes. Your data is encrypted while it is being sent and while it is stored in Looki's cloud. Access is strictly limited. When you use a cloud-based AI feature, Looki and its authorized AI providers may temporarily decrypt the content needed to provide that feature.
How long is media uploaded for cloud AI processing retained?
Raw media used only for temporary processing by Looki’s authorized AI providers is processed under the applicable zero-data-retention configuration. Content saved or synchronized to the user’s Looki account follows separate retention rules and may remain available until the user deletes it or closes the account.
0 comments