The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
For its documented capture-and-training workflow, Google Teachable Machine processes image, audio, and pose examples locally in your web browser. The product page says you can work entirely on-device without webcam or microphone data leaving your computer, and the legacy FAQ says that training happens locally rather than storing your images on Google servers. That protection applies to collecting examples and training the model. Exporting, hosting, embedding, or sharing the resulting model creates a separate privacy question because other people may be able to fetch its model files and metadata.
The short answer: local training, separate sharing risks
Teachable Machine is designed so that examples captured from a webcam, microphone, uploaded image files, uploaded audio files, or pose input are used in the browser. Google Creative Lab describes the tool as using TensorFlow.js to train and run models in the web browser. Its product page specifically says you can use it entirely on-device, without webcam or microphone data leaving your computer. The legacy Teachable Machine FAQ answers the storage question directly: “No. All the training is happening locally on your device.”
Those statements answer what happens during capture and training. They do not promise that every file, browser artifact, account event, or network request is private, and they do not make a published model private. When you export a model, the official image and audio examples show client code loading model.json and metadata.json from a model URL. Anyone who can access that URL may be able to retrieve those files and read labels or other metadata, depending on how and where you host them.
What “local” means in Teachable Machine
Local processing means the training computation runs in the browser on the device where you opened Teachable Machine. The browser receives the application code, then TensorFlow.js performs training with the examples available to that page. A webcam frame or microphone sample used for training is not documented as being uploaded to Google’s servers for that training step.
#1 Best Overall
- Privacy Protection and Lens Care: Avoid private information from hacking while preventing dust-fall and scratching of the camera lens
- Multiple Compatibility: Suitable for Logitech webcam C920x, C920, C922, C930e, C922x Pro Stream HD Camera
- Artful Design: Modeled and designed exclusively to fit the above devices from Logitech and make it more stylish
- Easy Flip Mechanism: Can be turned 180 angle and easily take the cover off when flipping more than 180
- Simple Installation: Attaches securely to your Logitech webcam without leaving residue, allowing for quick and hassle-free setup
| Stage | Where the documented work occurs | Privacy decision you still control |
|---|---|---|
| Choose or capture examples | In the browser, using files, a webcam, a microphone, or pose capture | Which people, rooms, documents, voices, and other content you expose to the device and page |
| Train the classifier | Locally in the browser with TensorFlow.js | Whether other software on the device can observe the screen, camera feed, microphone, or files |
| Test the model | In the browser against camera, microphone, pose, or uploaded input | Whether you keep permissions enabled after testing and whether sensitive test material remains on the device |
| Export the model | Model files are created for later use by client code | Where you host the files, who receives the URL, and whether the host provides access controls |
| Embed or share | A separate page or application fetches the exported files | Whether the URL is public, guessable, logged, cached, or shared beyond the intended audience |
Local training should therefore be read as a precise statement about the training path, not as a guarantee of zero data collection or absolute privacy for every action surrounding the tool.
How images are handled
Webcam images
Teachable Machine can capture image examples from a webcam after the browser grants camera permission. The documented on-device option means those camera examples can remain on the computer while the browser trains the image model. You should still check the browser’s camera indicator and stop capture when you are finished. A page that has permission may continue to access the camera until you revoke permission, close the page, or use the page’s stop control.
Uploaded image files
You can also supply image files instead of using a live camera. Selecting a file gives the page access to that file for the current workflow. Local training means the file is used by the browser rather than being documented as an upload for training on Google’s servers. Keep in mind that the original file may remain in your downloads folder, photo library, temporary files, backups, or browser history according to your operating system and browser settings.
What an image model may reveal
A trained classifier can expose information through its class names and behavior even when the original photos are not intentionally shared. Avoid labels such as a person’s full name, medical condition, home address, classroom roster, or other identifying description when you expect to publish the model. Use neutral labels when possible, and remove sensitive examples from the project and local storage when they are no longer needed.
Rank #2
- Privacy Protection: CloudValley webcam cover is designed for those who prioritize privacy, security, and peace of mind when using laptops, tablets, and computers
- Fashion Design: The space aluminum alloy webcam cover features a subtle design which compliments the beautiful aesthetic of top devices
- Ultra-Thin Design: Measures only 0.023 (0.6 mm) inch thin, ensuring it does not interfere with closing your laptop or device while providing reliable camera coverage
- Broad Compatibility: Works flawlessly with most laptops (MacBook, HP, Dell, Asus, Acer, Lenovo), All-in-One PCs and leading tablets including iPad, Surface Pro, Galaxy Tab, Fire HD, and Google Pixel Tablet
- Simple to Use: Only need to align to the webcam, attach and press it firmly for 15 seconds. Does not interfere with web use or indicator light
How audio and microphone data are handled
Microphone capture
The product supports recording sound examples with a microphone. The same local-processing statement applies: the documented workflow can train in the browser without microphone data leaving the computer. Watch for the browser’s microphone indicator, and revoke microphone permission after the session if you do not need it again.
Audio files
Audio files can be selected as training examples. Treat them like any other sensitive recording: a voice can identify a speaker, and background conversation can reveal people, locations, or confidential information. Local training does not erase the source file from the computer or prevent another program with access to that computer from copying it.
Audio export is a separate network event
The official audio integration snippet loads exported model files over HTTPS. That request concerns the model and its metadata, not a documented upload of your original microphone recordings during training. Once an audio model is hosted at a URL, however, the host and anyone allowed to fetch the URL become part of the privacy boundary.
Recommended Free Tools
Can Google see your webcam or microphone training data?
For the documented Teachable Machine training workflow, the product-specific answer is no: the homepage says webcam and microphone data can stay on the device, and the FAQ says training happens locally rather than storing images on Google servers. This is narrower than saying Google cannot receive any information from your browser under any circumstance.
Rank #3
- Note: Not suitable for MacBooks released after 2023 or devices with a protruding front camera; Not applicable to full-screen or notch-style tempered glass screen protectors; Do not use on the rear camera of the phone.
- 💻 Why Do You Need a Webcam Cover Slide? — Safeguard your privacy by covering your webcam with our reliable webcam cover when not in use. Don't let anyone secretly watch you. Stay protected!
- ✅ Thin & Stylish — Enhance your laptop's functionality and aesthetics with our 0.027" ultra-thin webcam covers. Seamlessly close your laptop while adding a touch of sophistication.
- ✅ Fits Most Devices — Compatible with laptops, phones, tablets, desktops! Keep your privacy intact on Ap/ple, Mac/Book, iPh/one, iP/ad, H/P, L/novo, De/ll, Ac/er, As/us, Sa/msung devices.
- ✅ 365 Days Protection — Our upgraded 3.0 adhesive ensures a strong hold that won't damage your equipment. Experience reliable, long-term privacy protection day in and day out.
The general Google Privacy Policy discusses media from interactions with Google services, account controls, and retention or deletion. It does not specifically establish that Teachable Machine’s local training samples are uploaded, nor does it provide a Teachable Machine-specific retention schedule for every browser-side artifact. Do not use the general policy as proof that local examples are sent to Google, and do not interpret the local-training statement as a promise that no telemetry, crash data, or other browser activity can ever exist.
What changes when you export or publish a model?
Exporting turns a private, local experiment into an asset that another page can load. The official examples show separate requests for model.json and metadata.json. The topology file describes the model structure; the metadata file can include class labels and other information needed by the client. The snippets demonstrate fetching those files from a model URL, but they do not promise authentication, private links, expiration, or access control.
Before publishing, assume that anyone who can obtain the model URL may be able to request the files. A public web host, a page source, browser developer tools, server logs, caches, and referrer data can all make a shared URL easier to discover. HTTPS protects the connection in transit; it does not make a publicly reachable file confidential.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Inspect an export before sharing
- Open the exported model location in a browser or download a copy to a controlled folder.
- Review
metadata.jsonfor class names, descriptions, and any labels that identify people, places, documents, or events. - Review
model.jsonand any companion files to confirm exactly what your host will serve. - Check the host’s sharing settings. Determine whether the URL requires authentication, expires, or is publicly readable.
- Test the URL from a separate browser session that is not signed in. If it loads there, treat it as public to anyone with the address.
- Remove the files from the host when the project ends, then check whether the host offers cache purging or deletion controls.
A privacy-focused Teachable Machine workflow
- Minimize the examples. Collect only the images or audio needed to distinguish your classes. Do not include bystanders, private screens, faces, full names, or background conversations unnecessarily.
- Use a controlled device. Keep the browser, operating system, and security software updated. Avoid collecting sensitive material on a shared or unmanaged computer.
- Grant permissions deliberately. Allow camera or microphone access only to the page that needs it. Check the browser indicator before and during capture.
- Prefer local files when appropriate. If you already have approved examples, use copies in a temporary project folder rather than exposing an entire photo or recording library.
- Train and test locally. Keep the project open only as long as needed, and stop the camera or microphone when testing ends.
- Sanitize labels. Use generic class names such as “class A” or “background sound” when a published model does not need descriptive or identifying labels.
- Separate private and public projects. Do not publish the same export that contains sensitive labels or examples used for internal testing. Create a clean export for public use.
- Control the model URL. Use a host with authentication, expiration, or access restrictions when the model is not intended for everyone. Do not paste a private URL into public source code.
- Clean up. Delete temporary source files, old exports, browser downloads, and hosted copies according to your organization’s retention rules.
Threats that local processing does not solve
- Browser extensions: An extension with access to pages, files, camera, or microphone content may observe activity independently of Teachable Machine’s training code.
- Malware or an untrusted account: Malicious software, remote-control tools, or another user with access to the device can capture the screen, files, camera feed, or microphone.
- Screen recording: Local training does not stop a screen recorder from saving the live preview, labels, or results.
- Operating-system storage: Source files, downloads, caches, backups, and crash artifacts may persist after a project closes. No Teachable Machine-specific retention period for these artifacts is documented.
- Shared URLs: An exported model can be copied, embedded, cached, or logged after you give someone its address.
A physical webcam cover can block the lens when the camera is not in use. It cannot mute a microphone, prevent browser network requests, stop an extension from reading a page, or secure an exported model.
Rank #4
- 【Premium Webcam Cover】This webcam privacy cover is an accessory of computer webcam. No worry about interfering with web camera lens use or indicator light; No damage to your device in any way as well. A helpful privacy protector and dust separator
- 【Privacy Protector】Slide the web camera cover over your webcam lens when not in use, and prevents web hackers from Spying on you. It is perfect to provide privacy security and peace of mind to individuals, groups, organizations, companies and governments. It also protects your camera lens from dust, and keeps it in high-definition resolution all the ways
- 【Durable Material】The web cam cover is made of high-strength plastic, which ensures that your privacy is protected for a long and lasting period of time. The back of the web camera privacy cover slide also has a strong 3M adhesive layer. It helps the privacy protector stick firmly to your device. The most convenient, super thin design, and extra mini size, make it perfectly combine with your devices
- 【Wide Compatibility】This webcam cover is compatible with most popular webcams with flat area surrounding lens or with protruding lens, such as Logitech HD Pro Webcam C920 C920x C930e and C922, Logitech C615 and C270 (NOT fit Logitech C910, B910, C310). It can be also used as a cover for the peep hole on door
- 【For Logitech Webcam Cover】 The streamcam cover kit comes with 2 pack. Please clean the lens surface before applying. Make sure the mounting surface is cleaned completely so that it sticks properly and firmly
Common privacy questions and practical answers
Does training upload my images?
The product-specific documentation says training happens locally and that on-device use can keep webcam or microphone data from leaving the computer. That statement covers the documented training workflow, not every other program or browser artifact on the device.
Does Teachable Machine upload my audio?
The documented microphone-training workflow is local in the browser. Treat the original recording, local copies, and any later hosted model as separate assets with their own access risks.
Can I make an exported model private?
Privacy depends on the service hosting the model files. Teachable Machine’s snippets show how a client fetches model and metadata URLs but do not promise private access controls. Use a host that supports authentication or restricted access if the model is confidential.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Are there guaranteed deletion dates?
No Teachable Machine-specific published schedule establishes how long local examples, browser caches, crash logs, or telemetry-related artifacts remain. Set and enforce your own deletion process, and verify deletion with the browser, operating system, and hosting provider you use.
Best Value
- 【Protect Privacy Security】Focusing on network security, now we can easily and effectively protect personal and family privacy security , Just gently slide the slide and close the camera, you can stop the intrusion of hackers.
- 【 Ultra Thin Design】The new ultra-thin design, with a thickness of only 0.022 inches, is made of flexible ABS material and is not fragile. Will not affect the closing of the laptops and scratch the laptops.
- 【Easy to install】 Strong adhesive makes the cover not fall, keep the screen clean and free of stains during installation, tear off the adhesive tape on the back, align it with our camera, and press hard for 10 seconds to work.
- 【Compatible with 】Compatible with camera for Laptop, tablet, computers, Echo Show and Apple Devices,as: MacBook Pro,Macbook Air,iMac ,Mac mini,iPad,MacBook Air, iPhone 6/7/8 Plus etc front camera .
- [What you get] 6 pack black webcam covers.
How to evaluate another browser-based machine-learning tool
If you are deciding whether a different tool is more suitable for sensitive work, compare documented behavior rather than assuming that every browser application is local. Ask these questions:
- Does training run locally, on a remote server, or in a choice of modes?
- Do webcam, microphone, and uploaded files leave the device?
- Are exported models and metadata publicly reachable by default?
- What account information, analytics, diagnostics, or telemetry are collected?
- Is there a written retention and deletion policy for source data, logs, and exports?
- Can administrators revoke access, expire links, and audit downloads?
For particularly sensitive recordings or images, verify these points with the tool’s own documentation and your organization’s privacy requirements before collecting data.
Bottom line
Teachable Machine’s documented image, audio, and pose training can run locally in the browser, so the product-specific evidence does not indicate that your training examples are automatically stored on Google servers. The privacy boundary changes when you export, host, embed, or share a model: its model files and metadata may be reachable through the URL, and the hosting service’s controls then matter. Use minimal examples, review permissions and labels, secure the export location, and avoid claiming “zero data collection” when the published documentation does not establish that broader guarantee.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
