When you use an AI chatbot, the service processes what you submit to produce a reply. Depending on the provider, account type, settings, region and features, that information may also be stored, reviewed for safety, used to improve models or services, or handled under a separate policy when you send feedback. A chat-history switch is not the same thing as a deletion command, and a model-training opt-out does not stop the processing needed to answer your prompt.
The practical rule is simple: treat every prompt as data moving through several systems, then check the controls for the exact chatbot and account you are using. OpenAI, Anthropic and Google document materially different defaults and exceptions, so there is no single privacy answer for “AI chatbots.”
What information can an AI chatbot receive?
The text you type is only one possible input. The exact categories depend on the product features you turn on, but a single request can combine several kinds of information.
Typed prompts and the reply
The service receives your prompt and generates a response from it. The prompt can contain personal details, business information, passwords, medical notes or confidential documents if you place those items in the text. The generated answer is also content associated with your session and may be covered by the provider’s storage, safety and improvement rules.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
#1 Best Overall
- HD streaming made simple: With America’s number 1 TV streaming platform,* exploring popular apps—plus tons of free movies, shows, and live TV—is as easy as it is fun. *Based on hours streamed—Hypothesis Group
- Compact without compromises: The sleek design of Roku Streaming Stick won’t block neighboring HDMI ports, and it even powers from your TV alone, plugging into the back and staying out of sight. No wall outlet, no extra cords, no clutter.
- No more juggling remotes: Power up your TV, adjust the volume, and control your Roku device with one remote. Use your voice to quickly search, play entertainment, and more.
- Shows on the go: Take your TV to-go when traveling—without needing to log into someone else’s device.
- TV, simplified: With setup that only takes minutes, a simple-to-navigate Home Screen, and an uncluttered remote control that does all you need—Roku makes it easier to watch the TV you love.
Files, images, video and audio
Uploading a document, photograph or video sends that material to the provider for processing. Voice features add an audio recording or a transcript. Some products retain an uploaded file in a separate library or storage area rather than treating it as part of one chat. That separation matters when you later delete a conversation.
Screen, browser and connected-app context
Features that read your screen, browser page, email, calendar or another connected service can send excerpts, summaries, generated media or inferences from those sources. Google’s Gemini notices expressly list screens, photos, imported chats, connected-app context, device and app information, and location information among data that may be processed, depending on the feature and settings. See the Gemini Apps Privacy Hub and its linked supplemental notice for the feature-specific scope.
Feedback and account information
A thumbs-up or thumbs-down report is not merely a private rating. The report can be accompanied by the conversation or recent context that explains why you sent it. Account, device and security information may also be processed to authenticate you, prevent abuse and operate the service.
Four different things providers may do with a prompt
Privacy settings are easier to understand when four separate operations are kept apart.
1. Process the request
The provider must handle your input, and often some surrounding context, to generate an answer. Turning off history or model improvement generally does not prevent this core processing. Safety filters may inspect the same material before or after the response is produced.
Rank #2
- Ultra-speedy streaming: Roku Ultra is 30% faster than any other Roku player, delivering a lightning-fast interface and apps that launch in a snap.
- Cinematic streaming: This TV streaming device brings the movie theater to your living room with spectacular 4K, HDR10+, and Dolby Vision picture alongside immersive Dolby Atmos audio.
- The ultimate Roku remote: The rechargeable Roku Voice Remote Pro offers backlit buttons, hands-free voice controls, and a lost remote finder.
- No more fumbling in the dark: See what you’re pressing with backlit buttons.
- Say goodbye to batteries: Keep your remote powered for months on a single charge.
2. Save a conversation or file
A chat may appear in your history, be archived, or be stored temporarily outside the visible history. An uploaded file can have its own storage record. “Not in history” therefore does not automatically mean “not retained anywhere.”
3. Review content for safety or quality
Automated systems can detect abuse, but providers may also route selected material to human reviewers. A conversation chosen for review can follow a different retention schedule from ordinary activity. Google states that even with Keep Activity off or when temporary chats are used, it still uses chats to respond and help protect Google, users and the public, including with help from human reviewers.
4. Use content to improve models or services
Training or service-improvement use is a future use of stored or selected content; it is not the same as the processing required to answer you. A control that prevents new conversations from improving a model may leave service delivery, security and legal retention unaffected. Feedback can create an exception to an otherwise restricted improvement setting.
How the documented services differ
The comparison below applies to the consumer products and account contexts described in each provider’s current documentation. It is not an industry-wide ranking, and the categories are not perfectly equivalent.
| Provider and surface | Model-improvement controls | History or temporary mode | Documented retention and deletion points | Feedback effect |
|---|---|---|---|---|
| ChatGPT individual services (Free, Plus and Pro) | “Improve the model for everyone” can be turned off in Settings > Data controls or through the Privacy Portal. New conversations after opting out are not used to improve models. | Temporary Chats do not appear in history, create or update memories, or improve models while temporary. | Temporary chats may be copied for safety for up to 30 days. Deleted chats are removed from the account view immediately and scheduled for permanent deletion within 30 days, subject to de-identification, security and legal exceptions. | Submitting response feedback may allow use of the entire associated conversation even after an individual user opts out. |
| Claude consumer offerings | Anthropic says chats and coding sessions may improve models when the user allows it, when content is flagged for safety review, or when the user otherwise explicitly opts in. Incognito chats are excluded from model improvement. | Incognito conversations are not used to improve Claude even when Model Improvement is enabled. | Deleted conversations disappear from history immediately and are automatically deleted from backend systems within 30 days, subject to other retention periods and legal or policy needs. | A thumbs-up/down report can store the entire related conversation, associated content and preferences for up to five years; Anthropic says feedback is de-linked from the user ID before use. |
| Gemini Apps | With Keep Activity on, Google says activity can provide, develop and improve services, including training generative AI models. With Keep Activity off, future chats are not used to train Google AI unless feedback is submitted. | Activity-off chats do not appear in Activity. Temporary chats are not used to train Google’s AI models, but both modes can still be processed for responses and safety. | Keep Activity uses an 18-month default auto-delete period, with choices of 3 months, 36 months or no auto-delete. Activity-off or temporary chats are retained for 72 hours. Chats reviewed by human reviewers and related data can remain for up to three years. | Feedback may include the feedback, the last 24 hours of chat context, and included uploads or connected-app content; reviewed feedback data can be kept for up to three years. |
Google’s Gemini Apps Privacy Hub was last updated September 24, 2026. Anthropic’s consumer guidance cited here is dated March 16, 2026. OpenAI’s consumer guidance was current when accessed October 13, 2026. Retention periods describe different categories, so they should not be treated as a cross-provider scorecard.
Rank #3
- 4K streaming made simple:With America’s number 1 TV streaming platform,* exploring popular apps—plus tons of free movies, shows, and live TV—is as easy as it is fun. *Based on hours streamed—Hypothesis Group
- 4K picture quality: With Roku Streaming Stick Plus, watch your favorites with brilliant 4K picture and vivid HDR color.
- Compact without compromises: Our sleek design won’t block neighboring HDMI ports, and it even powers from your TV alone, plugging into the back and staying out of sight. No wall outlet, no extra cords, no clutter.
- No more juggling remotes: Power up your TV, adjust the volume, and control your Roku device with one remote. Use your voice to quickly search, play entertainment, and more.
- Shows on the go: Take your TV to-go when traveling—without needing to log into someone else’s device.
OpenAI: settings, temporary chats and separate files
For individual ChatGPT services, the relevant control is Settings > Data controls > Improve the model for everyone. Turning it off applies to new conversations; it does not erase existing history or stop the service from processing prompts to answer them. The same choice is available through OpenAI’s consumer data overview.
Temporary Chat is a different mode. It keeps the conversation out of history, does not create or update memories and is not used to improve OpenAI models while it remains temporary. OpenAI may retain a safety copy for up to 30 days. If you save a Temporary Chat, it becomes a regular chat governed by ordinary settings.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchDeleting a regular or archived chat removes it from the account view immediately, but permanent deletion is scheduled within 30 days unless the content was already de-identified and disassociated or must be retained for security or legal reasons. Check for files in the Library separately: deleting the chat does not necessarily delete a file stored there.
OpenAI states that ChatGPT Business, Enterprise, Edu, Healthcare workspaces and the API are not used to improve models by default. An organization can choose to share data where that option is available. That named-workspace rule should not be generalized to every business account offered by every vendor.
Anthropic: opt-in improvement and feedback retention
Anthropic’s consumer guidance describes model improvement as conditional rather than automatic in every circumstance. A user can allow it; a conversation flagged for safety review can be used under the stated policy; and an explicit opt-in can authorize it. Incognito chats remain outside model improvement even if the account’s Model Improvement setting is enabled. The provider’s explanation is available in its model-training guidance.
Rank #4
- Stunning 4K and Dolby Vision streaming made simple: With America’s number 1 TV streaming platform,* exploring popular apps—plus tons of free movies, shows, and live TV—is as easy as it is fun. *Based on hours streamed—Hypothesis Group
- Breathtaking picture quality: Stunningly sharp 4K picture brings out rich detail in your entertainment with four times the resolution of HD. Watch as colors pop off your screen and enjoy lifelike clarity with Dolby Vision and HDR10+.
- Seamless streaming for any room: With Roku Streaming Stick 4K, watch your favorite entertainment on any TV in the house, even in rooms farther from your router thanks to the long-range Wi-Fi receiver.
- Shows on the go: Take your TV to-go when traveling—without needing to log into someone else’s device.
- Compact without compromises: Our sleek design won’t block neighboring HDMI ports, so you can switch from streaming to gaming with ease. Plus, it’s designed to stay hidden behind your TV, keeping wires neatly out of sight
The most consequential exception concerns feedback. Sending a thumbs-up or thumbs-down can cause the entire related conversation, including associated content and preferences, to be stored for up to five years. Anthropic says it de-links that feedback from the user ID before using it for service analysis, research, behavior study or model training as permitted by law. That five-year period is specific to feedback-associated consumer conversations, not a general statement about every Claude chat.
Recommended Free Tools
Deleting a conversation removes it from the visible history immediately. Anthropic says backend deletion occurs within 30 days, while noting that other retention periods and legal or policy requirements can apply.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Gemini: activity settings, reviewers and connected apps
Gemini’s central control is Keep Activity. When it is on, chats and shared content are saved in Gemini Apps Activity. Google says that activity can be used to provide, develop and improve services, including training generative AI models, and that human review may occur. The default auto-delete period is 18 months; users can choose 3 months, 36 months or no auto-delete.
Turning Keep Activity off changes future activity recording and training use, but it is not a zero-retention switch. Google says those chats are retained for 72 hours so the service can respond, process feedback and protect Google, users and the public. Temporary chats are not used to train Google’s AI models, yet Google says temporary and activity-off chats may still be used for responses and safety, including with human reviewers.
Google also describes a separate path for reviewed material: chats reviewed by human reviewers and related data can be retained for up to three years and are not deleted when ordinary activity is deleted. Feedback can include the report, the last 24 hours of conversation, uploads and connected-app content; reviewed feedback data can follow the same up-to-three-year period.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Best Value
- The Google TV Streamer (4K) delivers your favorite entertainment quickly, easily, and personalized to you[1,2]
- HDMI 2.1 cable required (sold separately)
- See movies and TV shows from all your services right from your home screen[2]; and find new things to watch with tailored recommendations for everyone in your home based on their interests and viewing habits
- Watch live TV and access over 800 free channels from Pluto TV, Tubi, and more[3]; if you find an interesting show or movie on your TV, mobile app, or Google search, you can easily add it to your watchlist, so it’s ready when you are[2]
- Up to 4K HDR with Dolby Vision delivers captivating, true-to-life detail[4]; and you can connect speakers that support Dolby Atmos for more immersive 3D sound
Connected apps add a second data relationship. Relevant excerpts, summaries, generated media or inferences can be used to answer a prompt and, with Keep Activity on, to improve services, including training. Disconnecting the app or deleting the original source data does not itself delete associated information already held in Gemini Apps Activity. Google’s privacy notice explains the import and connected-service cases.
Why deletion, history and training controls do not mean the same thing
Imagine uploading a spreadsheet, asking a question about it, receiving an answer and then deleting the visible chat. At least four records may have existed: the prompt and response, the uploaded file, safety or abuse-monitoring copies, and any feedback package you submitted. Each can have a different owner, control and clock.
- History controls determine whether a conversation is shown in your account activity.
- Training controls address specified future improvement uses; they do not necessarily stop service delivery or security processing.
- Deletion controls start a removal process but can have scheduled windows, legal exceptions, de-identification and separately stored files.
- Review controls concern selected content and may create a longer retention path than ordinary chats.
For that reason, “temporary,” “incognito” and “history off” should be read as limits on particular uses, not as promises that no copy exists.
A safer way to use an AI chatbot
- Identify the account context. Confirm whether you are in a personal account, a managed work or school workspace, an API project, or a shared family device. Policies can change with that context.
- Open the current data controls before entering sensitive material. Check the named provider’s settings, the account region and the date of the applicable notice.
- Minimize the prompt. Remove names, account numbers, exact addresses, passwords, private keys and any detail the chatbot does not need to answer the question. Replace real identifiers with fictional labels.
- Inspect uploads and connections. Ask whether a file is stored in a library, whether a connected app supplies excerpts, and whether disconnecting it will remove previously imported data. Do not assume it will.
- Choose the appropriate mode. Use temporary or incognito modes when you need their specific limits, but do not treat them as a guarantee of zero retention.
- Read the feedback prompt. Before selecting a rating, check what conversation window, uploads or preferences will be attached and whether feedback has a separate retention period.
- Clean up every storage location. Delete the conversation, then check libraries, activity pages, imported chats and connected-app records where the product provides separate controls.
- For regulated or confidential work, verify the contract and administrator settings. Do not infer enterprise protections from a consumer help page, or assume one vendor’s business rule applies to another.
Questions to ask before trusting a chatbot with confidential data
- What exact data does this feature send: text only, or also files, voice, screen, location and connected-app content?
- Is model improvement on by default for this account, and does opting out apply only to new conversations?
- Can human reviewers see selected chats, and what retention period applies to reviewed material?
- Where are uploaded files stored, and are they deleted separately from the chat?
- What happens if I submit feedback?
- Which deletion exceptions apply for safety, legal obligations, security or de-identification?
These questions produce a more accurate answer than labeling a chatbot simply “private” or “not private.” The meaningful choice is whether the data flow, retention path and improvement setting fit the information you are about to submit.
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.
Say something about this film