To keep code private, check the full data path—not just where the model runs. A local model can reduce exposure to a model provider, but an editor extension, agent tool, telemetry service, or session sync could still send code elsewhere. A cloud service may offer strong contractual and administrative protections, but you need to check the exact product, plan, model, settings, and retention terms. Treat training and storage as separate questions: a provider may not train on your prompts while still retaining them for other purposes.
What “private” means when an AI assistant handles code
Code privacy is not a single setting. An assistant may receive a prompt, selected snippets, repository context, tool results, or a conversation history. Those data can travel through more than one service: the editor or extension, an agent runtime, a model host, telemetry, and features that sync sessions. Identify those routes before deciding whether a workflow is acceptable for your code.
- Transmission: Does code stay on the device, or is any of it sent to a provider or another service?
- Training: May prompts, outputs, or other interactions be used to train or improve models? Is that controlled by a default, an opt-in, or an account setting?
- Retention: What is kept, for how long, and for which purpose? Prompts, generated responses, abuse-monitoring logs, feedback conversations, and application state can have different rules.
- Access and scope: What repository files and tools can the assistant access, and which employees, administrators, or service providers can review data under the applicable terms?
- Contract and geography: Which product agreement, organization controls, model host, and processing region apply?
A “no training” commitment answers only the training question. It does not, by itself, establish that data is never logged, retained, reviewed, or synchronized.
Is a local AI model private?
Local inference can keep the model’s processing on your device, but “local” is a configuration property, not a blanket privacy guarantee. The answer depends on the whole setup: the runtime, editor integration, plugins, agent tools, telemetry, and any synchronization or backup features. The sources cited here do not certify a particular local application or configuration.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →#1 Best Overall
- 【Instant Snap-on Magnetic Attachment】- The Patented Magnetic Privacy Screen – Protected by U.S. Patents 9,829,669 and D844,012. Simply place the privacy screen along the top of your MacBook and let the magnets attach along the top. No need for tricky placement, messy tape, or damaging adhesive. Easily remove and reattach when you need it.
- 【Filter Dimensions】: Width: 11 15/16" (304 mm), Height: 7 1/2" (190 mm), Diagonal: 14.1" (358.14 mm) - SightPro Blackout Privacy Filter is engineered to be compatible with Lenovo, HP, Dell, Acer, Asus, Samsung, and other laptop brands. Please verify your screen's width and height measurements before ordering. It's not recommended to make your selection based solely on your screen's diagonal size. [Not optimized for touchscreens.]
- 【Superior Privacy】- Our advanced multi-layered film filter blacks out your screen when viewing from the side, while maintaining a crystal clear screen straight-on. It also protects your eyes from harmful UV and blue light. [Note: It does not block visibility directly behind you, regardless of the distance.]
- 【Perfect for Travel and Open Workspaces】- The Laptop Privacy Screen Filter is the ideal solution for healthcare providers, mobile workers, commuters, students, and business travelers. Now you can stay compliant and safeguard sensitive corporate information while working in airplanes, subways, airports, and public areas.
- 【Package Contents】- Each package includes a magnetic privacy screen filter, magnetic stickers, a webcam privacy cover, a storage folder, and a cleaning cloth. Buy with confidence – located in the US, Sight Pro specializes in providing best-in-class privacy solutions to individuals, small businesses, corporations, government, and educational institutions. Our privacy screens are Section 889 and TAA compliant.
Check the complete local workflow
- Confirm that the selected model is actually running on-device rather than routed to a hosted endpoint.
- Review the editor extension and runtime’s documentation and settings for cloud requests, telemetry, crash reports, and session sync.
- Check whether agent tools can call external services, browse the web, or send file contents outside the local runtime.
- Limit repository access to the files needed for the task; do not assume that a local model needs or receives the whole repository.
- Verify network behavior using controls approved by your organization. Do not infer isolation solely from a “local” label.
Local inference may suit work that must not be sent to an external model provider, but it does not automatically prevent other components in the workflow from transmitting code.
Does AI use my code for training, and does a no-training policy mean it is not stored?
Those are separate policy questions, and the answer depends on the exact service and account type. For example, OpenAI’s API data-controls documentation says API data is not used to train or improve models unless the customer opts in. The same documentation says default abuse-monitoring logs may include prompts, responses, and derived metadata, and may be retained for up to 30 days, subject to stated legal and safety exceptions. Eligible customers may apply for Modified Abuse Monitoring or Zero Data Retention; eligibility, approval, feature limits, and exceptions apply. This 30-day figure is for the stated default API abuse-monitoring logs, not every product or all application state.
Rank #2
- Compatible Models: Width: 13 9/16" (13.5 inch/344 mm), Height: 7 5/8" (7.6 inch/194 mm), Diagonal: 15.6" (396.24 mm) widescreen laptops which have a 16:9 aspect ratio. Not touchscreen compatible !!! Not fit for 16:10.Do NOT rely solely on your laptop’s diagonal size when ordering. Use a ruler to measure your screen’s visible area (excluding the black bezels). If the width reads 344mm and height reads 194mm, this filter is a perfect match for your device.
- Keep Information Privacy: Effective "black out" privacy from side views outside the 60-degree viewing angle. Designed for optical clarity when viewing from the front, a person not at the front of the screen can only see the dark side of the screen, so it protects buisness secrets and personal privacy
- Eye and Screen Protection: Privacy filter does not only protect your private life but also protects your eyes by blocking 30% of blue light , blocking the harmful blue light between 380 - 495nm, it filters out the blue light and relieves eye strain. Our laptop privacy screen also helps keep your screen safe from dust and scratches
- Perfect For Open Workspaces: Great for maintaining screen privacy in high traffic areas such as open work spaces, airports, airplanes, commuter trains, coffee shops and other public places, etc
- Easy Installation: Choose between 2 simple Options; Slide-On/Off or Mounted. Not touchscreen compatible
OpenAI separately says that ChatGPT Enterprise, Business, Edu, Healthcare, Teachers, and the API platform do not use inputs or outputs for training by default. Its business-data page also describes encryption in transit and at rest and retention controls for qualifying organizations. These commitments are scoped to the listed business products and API platform; they should not be generalized to unnamed consumer products or other providers. OpenAI’s stated commitment is: “We don’t train our models on your organization’s data by default.”
Anthropic’s consumer privacy article, dated March 16, 2026, covers Free, Pro, Max, and those accounts using Claude Code. It says consumer chats and coding sessions may be used to improve models if the user opts in, if a conversation is flagged for safety review, or through another explicit opt-in. It also says feedback can result in the related conversation being stored for up to five years. That five-year period applies to the related conversation stored with feedback under the described consumer policy; it is not a general retention period for all Claude use. The article says, “Your Incognito chats are not used to improve Claude, even if you have enabled Model Improvement in your Privacy Settings.” Commercial and API users should consult Anthropic’s separate information rather than assume the consumer policy applies.
Free tools Windows power users keep installed
One-click scans. No signup required.
Rank #3
- Filter Dimensions: Width: 11 15/16" (304 mm), Height: 7 1/2" (190 mm), Diagonal: 14.1" (358.14 mm) - SightPro Blackout Privacy Filter is engineered to be compatible with Lenovo, HP, Dell, Acer, Asus, Samsung, and other laptop brands. Please verify your screen's width and height measurements before ordering. It's not recommended to make your selection based solely on your screen's diagonal size. [Not optimized for touchscreens.]
- Two Attachment Options - Installs in minutes. Option 1 uses clear adhesive strips that securely attach to any screen. Option 2 uses slide mount tabs that easily stick to the display frame, allowing you to slide the filter on and off the screen as needed.
- Superior Privacy and Anti Glare - Our advanced multi-layered film filter blacks out your screen when viewing from the side, while maintaining a crystal clear screen straight-on. It also protects your eyes from harmful glare, UV, and blue light. [Note: It does not block visibility directly behind you, regardless of the distance.]
- Perfect for Travel and Open Workspaces - Our computer screen privacy filter is the ideal solution for healthcare providers, mobile workers, commuters, students, and business travelers. Now you can stay compliant and safeguard sensitive corporate information while working in airplanes, subways, airports and public areas.
- Package Contents - Each package includes one privacy screen shield filter, two sets of clear adhesive strips, two sets of slide mount tabs, and a microfiber cleaning cloth. Buy with confidence – located in the US, Sight Pro specializes in providing best-in-class privacy solutions to individuals, small businesses, corporations, government, and educational institutions. Our privacy screens are Section 889 and TAA compliant.
What code does a cloud coding assistant send?
The amount and destination of context can vary by product, feature, model, and account. A chat prompt might include only text and a selected snippet; an agent can also work with repository files and tool outputs. Check the client’s context and permissions rather than assuming that only the words typed into the prompt are transmitted.
GitHub Copilot model hosting and account scope
GitHub documents different model hosts and data commitments by model. Its documentation says Copilot Business and Enterprise data is not used by GitHub to train AI models. For individual subscribers, interaction data—including prompts, suggestions, and generated snippets—may be used for training under GitHub’s privacy statement and settings; GitHub describes an opt-out setting. The same documentation notes model-specific conditions and exceptions, including retention for certain models and a time-bounded Zero Data Retention exception. Because the model list and conditions can change, check the current entry for the specific model and account before enabling it.
Rank #4
- 【Instant Snap-on Magnetic Attachment】- The Patented Magnetic Privacy Screen – Protected by U.S. Patents 9,829,669 and D844,012. Simply place the privacy screen along the top of your MacBook and let the magnets attach along the top. No need for tricky placement, messy tape, or damaging adhesive. Easily remove and reattach when you need it.
- 【Filter Dimensions】: Width: 12 3/16" (310 mm), Height: 6 7/8" (175 mm), Diagonal: 14" (355.6 mm) - There are two different 14 inch screen sizes, please select the correct one. SightPro Blackout Privacy Filter is engineered to be compatible with Lenovo, HP, Dell, Acer, Asus, Samsung, and other laptop brands. Please verify your screen's width and height measurements before ordering. It's not recommended to make your selection based solely on your screen's diagonal size. [Not optimized for touchscreens.]
- 【Superior Privacy】- Our advanced multi-layered film filter blacks out your screen when viewing from the side, while maintaining a crystal clear screen straight-on. It also protects your eyes from harmful UV and blue light. [Note: It does not block visibility directly behind you, regardless of the distance.]
- 【Perfect for Travel and Open Workspaces】- The Laptop Privacy Screen Filter is the ideal solution for healthcare providers, mobile workers, commuters, students, and business travelers. Now you can stay compliant and safeguard sensitive corporate information while working in airplanes, subways, airports, and public areas.
- 【Package Contents】- Each package includes a magnetic privacy screen filter, magnetic stickers, a webcam privacy cover, a storage folder, and a cleaning cloth. Buy with confidence – located in the US, Sight Pro specializes in providing best-in-class privacy solutions to individuals, small businesses, corporations, government, and educational institutions. Our privacy screens are Section 889 and TAA compliant.
What BYOK changes—and what it may not change
GitHub says that when Copilot Chat uses bring your own key (BYOK), prompts and responses are transmitted to the selected provider and may be subject to that provider’s retention and privacy policies. BYOK therefore changes the destination for those prompts and responses; it does not necessarily route every agent action through that provider. GitHub cautions that some code-application or tool actions in Agent mode may still use Copilot-integrated models. Check the feature-specific route, not just the key configuration.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Compare the actual data path before choosing a workflow
Use this as a comparison framework, not a ranking. The policies below describe only the named services and scopes; they do not establish what an individual editor, extension, or organization has configured.
Best Value
- 【Instant Snap-on Magnetic Attachment】- The Patented Magnetic Privacy Screen – Protected by U.S. Patents 9,829,669 and D844,012. Simply place the privacy screen along the top of your MacBook and let the magnets attach along the top. No need for tricky placement, messy tape, or damaging adhesive. Easily remove and reattach when you need it.
- 【Filter Dimensions】: Width: 13.56" (344.5 mm), Height: 8.49" (215.6 mm), Diagonal: 16" (406 mm) - SightPro Blackout Privacy Filter is engineered to be compatible with Lenovo, HP, Dell, Acer, Asus, Samsung, and other laptop brands. Please verify your screen's width and height measurements before ordering. It's not recommended to make your selection based solely on your screen's diagonal size. [Not optimized for touchscreens.]
- 【Superior Privacy】- Our advanced multi-layered film filter blacks out your screen when viewing from the side, while maintaining a crystal clear screen straight-on. It also protects your eyes from harmful UV and blue light. [Note: It does not block visibility directly behind you, regardless of the distance.]
- 【Perfect for Travel and Open Workspaces】- The Laptop Privacy Screen Filter is the ideal solution for healthcare providers, mobile workers, commuters, students, and business travelers. Now you can stay compliant and safeguard sensitive corporate information while working in airplanes, subways, airports, and public areas.
- 【Package Contents】- Each package includes a magnetic privacy screen filter, magnetic stickers, a webcam privacy cover, a storage folder, and a cleaning cloth. Buy with confidence – located in the US, Sight Pro specializes in providing best-in-class privacy solutions to individuals, small businesses, corporations, government, and educational institutions. Our privacy screens are Section 889 and TAA compliant.
| Workflow or policy scope | Training use stated in the cited documentation | Retention or routing detail | What to verify |
|---|---|---|---|
| Local inference | Not established for a particular local setup by the sources cited here. | Depends on runtime, editor, extensions, tools, telemetry, and synchronization. | Confirm on-device inference and inspect every connected component and its network behavior. |
| OpenAI API | API data is not used to train or improve models unless the customer opts in, according to OpenAI’s API data-controls documentation. | Default abuse-monitoring logs may contain prompts, responses, and derived metadata and may be retained for up to 30 days, subject to stated exceptions. Eligible customers may seek Modified Abuse Monitoring or Zero Data Retention. | Check eligibility, approval, feature limitations, exceptions, and whether the application stores additional state. |
| OpenAI listed business products and API platform | Inputs and outputs are not used for training by default for ChatGPT Enterprise, Business, Edu, Healthcare, Teachers, and the API platform, according to OpenAI’s business-data page. | The page describes encryption and retention controls for qualifying organizations; a universal retention period is not stated there. | Confirm the named product, qualifying organization controls, applicable agreement, and retention configuration. |
| GitHub Copilot | GitHub says Business and Enterprise data is not used by GitHub to train AI models. Individual subscribers’ interactions may be used under GitHub’s privacy statement and settings; an opt-out setting is described. | Hosting, retention, and exceptions vary by model. With BYOK, prompts and responses go to the selected provider; some Agent mode actions may still use Copilot-integrated models. | Check the current model entry, account type, settings, provider terms, and route for each agent action. |
| Anthropic consumer products | For Free, Pro, Max, and consumer accounts using Claude Code, chats and coding sessions may be used to improve models in the stated opt-in and safety-review circumstances. | Feedback can result in the related conversation being stored for up to five years. This is not a general retention period for all Claude use. | Confirm whether the account is consumer or commercial/API, and review the applicable settings and policy. |
The OpenAI API’s up-to-30-day default abuse-monitoring period and Anthropic’s up-to-five-year feedback-related period describe different data and purposes; they are not directly comparable retention measures.
A practical checklist for keeping code private
- Classify the code first. Follow your organization’s source-code and data-handling rules. Do not paste confidential or restricted code into a tool that has not been approved for it.
- Map the route for each feature. Identify the editor, extension, model runtime or host, BYOK provider, agent tools, and any session synchronization. A single product may use more than one route.
- Read the terms for the exact configuration. Record the product, plan, model, account type, geography, training settings, retention rules, feedback handling, abuse monitoring, and any human-review provisions or exceptions.
- Send less context. Remove credentials and secrets, use the smallest relevant snippet, and avoid granting broad repository context unless the task requires it. Restrict agent tools and file permissions to what is necessary.
- For cloud use, use an approved organization workflow. Confirm the applicable agreement and administrative controls rather than relying on a public no-training statement alone.
- For local use, check the surrounding software. Verify the runtime, extension, telemetry, external tool calls, and sync settings instead of treating “local” as proof that nothing leaves the machine.
- Recheck after changes. Revisit settings and current terms when the model, plan, product features, or account configuration changes. Review generated code before using it.
Sources and policy freshness
The provider-specific statements above reflect the official OpenAI API data-controls and business-data pages, GitHub Copilot model-hosting and responsible-use documentation, and Anthropic’s consumer privacy article dated March 16, 2026. These policies, model lists, and exceptions can change; verify the current terms and settings for the exact product and model you plan to use.
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.
