Recommended Free Tools
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
An AI chatbot collects a child’s conversation by receiving what the child types, says, or uploads, and it may also require account details such as age, a parent’s contact information, or billing details for a paid app. The service processes those inputs to generate a reply. Storage, review, sharing, model improvement, and deletion then follow that provider’s privacy notice, settings, age rules, and the applicable law. There is no single retention period and no single model-training rule for every chatbot. In orders issued September 10, 2025, the U.S. Federal Trade Commission asked seven companion-chatbot companies how they handle personal information from conversations. The orders ask for information. Public findings that every chatbot sells or permanently keeps children’s chats are outside those documents.
What can show up inside a child’s chatbot conversation?
The conversation itself can be personal information, including details a child never entered in a profile. A thread can hold health worries, feelings, friendship problems, family circumstances, and identifying context such as a name, a school, a neighborhood, or a daily routine.
UNICEF’s 2025 Guidance on AI and Children, version 3, cautions that children may reveal sensitive information when they use AI to talk about emotions, friendships, or health, and that they may not appreciate the sensitivity of what they typed. Ordinary use is enough to create the risk. A child who would hesitate to post a medical fact on a public page may still type it into a thread that looks private and answers in a friendly voice.
A constructed example shows how the categories stack inside one message. A child writes: “I can’t sleep because I’m worried about my asthma check at the clinic near my school, and I don’t want my friends to know.” That single text contains a health condition, a care appointment, a rough location, a link to a school, and a social worry. Nobody asked for those items on a form. They are now inputs the service processes in order to reply. If the product accepts a photo of a prescription label, or a voice note that says the same thing, that file sits in the same problem whenever the service receives it.
Popular on Amazon
- 【90 Degree Full Function USB C Adapter】Supports the latest USB4 standard for up to 40Gbps of bandwidth, 8K video support, and up to 240W Fast charging; Support USB PD 3.1 Extended Power Range (EPR) including 140W, 180W and 240W. Your productivity is easily doubled. *Warm Tip: For the best fit, please make sure the USB-C port is fully exposed; thick or bulky cases may affect compatibility.
- 【Reliable Quality】Made from high-quality, scratch-resistant aluminum alloy, delicate metal frame. get excellent wear resistance and heat dissipation. Non-Slip Design for easier insertion and removal.
- 【Small and portable】Ultra thin size, ultra compact design, a petite physique that matches the minimalist design of your modern laptops. Coming with a slim figure that won't block the port next to the one it's plugged into.
- 【Compatibility】Widely work with most usb c, thunderbolt 4/3 devices. Compatibilit for ROG Ally, Steam Deck, Switch, MacBook Pro, Quest 3 VR headset, Surface Book 2, iPhone 18/17/16/15, Samsung Galaxy S23/S22/S21, Note 20/10, Google Pixel, ChromeBook, etc.
- 【Excellence Customer Support】If you experience any issues or cable failure, simply reach out to our dedicated customer service team. We provide a prompt response within 18 hours, providing you with the necessary technical support to resolve any concerns.
Companion-style design changes how much children are willing to say. The Federal Trade Commission’s Generative AI 6(b) resolution, issued September 10, 2025, states the concern this way:
And because AI companions are designed to communicate like a friend or trusted confidant, children may be more inclined to share sensitive personal information, which companies may then sell to third parties or exploit in an unexpected manner, further undermining children’s privacy interests or increasing the risk of related adverse real-world effects.
That passage is the Commission’s rationale for an inquiry into a risk. It is the Commission’s language for why the design matters. Individual findings that a named company sold children’s information are not what this sentence reports.
Version 3 of the UNICEF guidance discusses the same pattern as relational AI: systems that invite an ongoing, personal-feeling relationship call for child-safety and privacy protections that a one-off factual answer does not. The guidance says protecting children’s data includes using it only with informed consent and only lawfully, explaining practices in age-appropriate language, and refusing indiscriminate collection. Collection should be proportionate to a stated purpose. Those are child-rights standards in UNICEF’s text. They are a checklist for reading a product. The statute or regulation that binds a specific company remains the legal duty.
What can the service collect before a child sends the first message?
Account screens often collect personal information that never appears in the transcript. UNICEF notes that some systems require basic details such as a child’s age, a parent’s contact information, or billing details when the app is paid. The exact fields vary. One product may open without an account. Another may wait for a birth date and a parent email. A paid upgrade may add a payment method the child never mentions in the chat.
Account data and conversation data are different collections. A control that hides yesterday’s thread leaves the question of a stored parent email or payment method to the account and billing sections of the notice. Families who look only at the chat window are looking at one of the piles.
A third category sits beside those two. A service may also hold operational or inferred information tied to the session: material the child did not type as a message and did not enter on the signup form. The FTC, UNICEF, ICO, and European Commission sources drawn on here do not publish a complete technical inventory of that category for every chatbot. They do establish conversational disclosures and the account-detail examples above. For logs, identifiers, or profiles beyond those, the service’s current privacy notice is the page that has to describe them.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →| Category | What the cited sources support | Where to look |
|---|---|---|
| Conversation content | Text the child sends; sensitive topics UNICEF names (emotions, friendships, health); and, where the product accepts them, uploads, voice inputs, and transcripts | The sections of the notice on messages, files, audio, retention, review, training, and deletion |
| Account and payment data | Age, a parent’s contact information, and billing details for paid apps. Requirements vary by service | Signup fields plus the account and payment sections of the notice |
| Other operational or inferred data | A possible further category. These official sources do not itemize it for every chatbot | Any notice language about logs, identifiers, or information the company says it derives |
Before a child agrees, write down every signup field. Age, parent contact, and payment data are the examples UNICEF gives. If the screen also asks for a phone number, a school, or a photo, treat that as further account data and find the sentence that states its purpose. Purpose-specific, minimal collection is the UNICEF benchmark. A required field with no stated purpose is a reason to decline the account or to ask the company why the field exists.
What happens to a message after a child sends it?
Two steps are common to the chatbots this subject covers. The service receives the child’s content, and it processes that content to produce a response. Retention, human review, sharing, product improvement, model training, and deletion are the next decisions, and each one follows that company’s current policy, settings, and legal duties.
The Federal Trade Commission treated those later decisions as questions companies should answer. On September 10, 2025, the FTC announced orders to seven companies that provide consumer-facing AI chatbots acting as companions. The orders seek information on monetization of engagement, how user inputs are processed, safety testing, disclosures, age rules, and the use or sharing of personal information from conversations. The press release says the inquiry also seeks information on “how these firms measure, test, and monitor potentially negative impacts of this technology on children and teens.”
Also popular on Amazon
- 5-in-1 Connectivity: Equipped with a 4K HDMI port, a 5 Gbps USB-C data port, two 5 Gbps USB-A ports, and a USB C 100W PD-IN port. Note: The USB C 100W PD-IN port supports only charging and does not support data transfer devices such as headphones or speakers.
- Powerful Pass-Through Charging: Supports up to 85W pass-through charging so you can power up your laptop while you use the hub. Note: Pass-through charging requires a charger (not included). Note: To achieve full power for iPad, we recommend using a 45W wall charger.
- Transfer Files in Seconds: Move files to and from your laptop at speeds of up to 5 Gbps via the USB-C and USB-A data ports. Note: The USB C 5Gbps Data port does not support video output.
- HD Display: Connect to the HDMI port to stream or mirror content to an external monitor in resolutions of up to 4K@30Hz. Note: The USB-C ports do not support video output.
- What You Get: Anker 332 USB-C Hub (5-in-1), welcome guide, our worry-free 18-month warranty, and friendly customer service.
That list maps what the Commission considered material in September 2025. It is an investigative request. The press release and the authorizing resolution, in the passages used here, do not state completed findings against a named company, and they do not publish one retention schedule or one training practice for all seven firms. Later case developments, if any, are outside these two documents. A summary that collapses the inquiry into “chatbots sell children’s chats” adds a conclusion the September 2025 texts do not report.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsFor a household, the same list is a reading guide for the privacy notice:
- Processing. The notice should describe what happens to text, and to voice or uploads if the product has them, in order to generate answers.
- Use and sharing. The notice should say whether conversation data is used for the reply, for safety review, for monetization, for product improvement, or for training, and who else receives it.
- Age and disclosure. The notice, or a child-safety page it links to, should say who the service is for and what parents are told.
- Impacts on children and teens. A company asked how it measures effects on children and teens may describe that work in public. Silence in the notice is a gap in what a family can check from the outside.
A service can process a message only far enough to answer and keep a narrow policy. It can also keep inputs for further purposes. The current notice and the settings on the account the child uses are the places the company says which of those descriptions applies.
Do AI chatbots save what kids say?
Some services store prompts, uploads, voice inputs, or transcripts. Whether a particular service does, and for how long, is a fact in that service’s privacy notice and account settings. The FTC, UNICEF, ICO, and European Commission materials cited here do not establish a retention period shared by chatbots as a group. A claim that “kids’ chats are kept for a set number of days” needs the notice of the app in front of you, dated on the day you read it.
Read four statements separately. Notices often fold them into one paragraph.
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 match- Whether content is stored after the reply. Look for a description of keeping the conversation, or of treating the exchange as short-lived. If the child could open more than one plan or product, read the notice that matches the one on the device. These sources do not establish that every plan from a company uses the same retention rule.
- Which kinds of content the storage rule covers. Text, files, images, and voice can be described in different sections. A paragraph about “messages” leaves the microphone question open. If the child uses voice, find the audio section before applying the text rule to recordings.
- Who can review what is stored. The notice should say whether staff or service providers can read conversations, and for the purposes it names. Storage and review are separate events. A company can describe retention with a review program, or retention with no routine human reading. The FTC’s questions on use and sharing exist because those designs differ.
- How long storage lasts, and what deletion covers. Look for a period, for the event that starts it, and for exceptions. Then read the deletion control beside that language. Nothing in the cited official sources shows that a control labeled delete, clear history, or similar removes every copy a company or its service providers hold. If the notice mentions backups, security records, or data already used to improve a model, those limits belong in your notes before you rely on the button.
Confirm you are reading the settings for the account the child actually uses. Notices and controls are described per service and per account. A policy page for a different product, or a help article from a previous year, can describe a rule the installed app no longer uses.
Can chatbot conversations be used to train AI?
A provider may use conversation data to improve a product or train a model when its terms say so and the law that applies allows it. The FTC’s September 2025 orders ask how companies process inputs and how they use or share personal information from conversations. Those orders do not establish that every companion chatbot trains on children’s chats, and they do not establish that none of them do. Training and product improvement are uses to look up in the specific notice.
In the privacy notice, search for “train,” “improve our models,” “develop our services,” and “human review.” On the settings screen, search for a control that limits any of those uses. If a control exists, read what it claims to change going forward. The UNICEF, FTC, ICO, and European Commission materials cited here do not describe a switch that removes information from a model already trained, or that deletes every copy a service provider holds. A setting can govern later chats and leave an earlier use where it is. Where the company draws that line, the notice is the place it will say so.
UNICEF’s version 3 guidance is a standard for judging the language you find. Children’s data should be collected for a purpose, limited to what that purpose needs, and used lawfully and with informed consent. The guidance warns against indiscriminate collection. A notice that allows chats to be used for a wide list of unnamed future purposes sits poorly against that standard. A notice that names a narrow purpose, explains it in language a child can follow, and offers a real choice sits closer to it. Closeness to the UNICEF standard is a child-rights evaluation. Legal compliance is a separate conclusion under the law that covers that operator.
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 →If the notice never mentions training, leave the blank as a blank. Ask the company whether conversation content from children’s accounts is used to train or improve models, and keep the reply with the notice. Until an answer exists, the step that matches data minimization is to limit what the child sends. Silence in a policy is an unanswered question.
Also popular on Amazon
- Sleek 7-in-1 USB-C Hub: Features an HDMI port, two USB-A 3.0 ports, and a USB-C data port, each providing 5Gbps transfer speeds. It also includes a USB-C PD input port for charging up to 100W and dual SD and TF card slots, all in a compact design.
- Flawless 4K@60Hz Video with HDMI: Delivers exceptional clarity and smoothness with its 4K@60Hz HDMI port, making it ideal for high-definition presentations and entertainment. (Note: Only the HDMI port supports video projection; the USB-C port is for data transfer only.)
- Double Up on Efficiency: The two USB-A 3.0 ports and a USB-C port support a fast 5Gbps data rate, significantly boosting your transfer speeds and improving productivity.
- Fast and Reliable 85W Charging: Offers high-capacity, speedy charging for laptops up to 85W, so you spend less time tethered to an outlet and more time being productive.
- What You Get: Anker USB-C Hub (7-in-1), welcome guide, 18-month warranty, and our friendly customer service.
Why do companion-style chatbots raise a sharper privacy risk?
Companion chatbots are built to keep talking and to sound like a friend. The FTC’s September 10, 2025 resolution treats that design as a reason children may share sensitive personal information more readily than they would with a tool that only answers a question. The resolution then identifies a further risk: companies may sell that information to third parties or exploit it in an unexpected way, with effects on privacy and on life offline. The document authorizes investigation of that risk. Completed findings about a particular firm are not in the resolution text relied on here.
The order topics show what the Commission wanted explained alongside that risk. Monetization of engagement appears beside processing, sharing, safety testing, disclosures, and age rules. A companion optimized to hold attention produces more conversation, and more conversation can produce more personal detail. UNICEF’s version 3 checklist and the full guidance approach the same structure from the child-rights side: a relationship-like system needs privacy protections and safety protections, not only fluent replies.
Two household consequences follow.
The first is about the moment a child decides to disclose. A child who treats the bot as a diary will put diary-level facts into software run by a company. The explanation to give before that habit starts is that a warm reply still goes to the service that operates the bot. Tone is a design feature. It does not change who receives the text.
The second is about time. Companion products are built for return visits. One careful message and months of nightly chats about school, friends, and health are different bodies of data. Retention and training terms matter more as the history grows. Reading those terms before the account exists is easier than reconstructing them after a long thread.
The Commission’s press release also asks how firms measure negative impacts on children and teens. The inquiry’s child-safety questions reach teenagers, including teens who sit above the age line of a particular children’s privacy statute. A published minimum age of 13, or of 16, can coexist with a serious question about how teen conversations are used. Age gates and children’s privacy statutes answer different questions. The next section keeps those rules in their own countries.
Which privacy rules apply to children’s chatbot data?
The duties that matter are the ones that cover the operator and the child. Three official sources are easy to blur together, and they should stay labeled. An accurate account keeps each duty inside its own legal setting.
United States. The FTC’s Complying with COPPA: Frequently Asked Questions explains requirements for operators covered by the Children’s Online Privacy Protection Rule. Where the rule applies, those requirements include direct notice to parents about covered collection, use, or disclosure of children’s personal information, and parental consent. COPPA is a U.S. rule for covered operators and covered circumstances. Using AI does not place every chatbot and every teenager under the same duties. Whether a given operator is covered depends on the tests the rule and the FAQ describe. Parents can use the FAQ to see what covered operators must tell them and when consent is required. The FAQ is the business-facing explanation of that U.S. framework.
Free tools Windows power users keep installed
One-click scans. No signup required.
United Kingdom. The Information Commissioner’s Office explains common data-processing activities that can affect children’s rights, including privacy, informed choices, transparency, and data minimization. For a child in the UK, those are practical lenses on a chatbot notice: whether the child and the parent can tell what is collected, whether collection is limited, and how the processing affects the child’s rights. The page is written for organizations assessing their own practices. Families can use its list of impacts as a reading guide. It belongs to the UK children’s-information materials.
European Union. In 2025 the European Commission published guidelines on measures to ensure a high level of privacy, safety, and security for minors online. The guidelines say AI chatbots and companions may affect minors’ privacy, safety, and security. They describe age assurance as one possible way to reduce exposure to risks. In that document, age assurance is a measure about who reaches a service or a feature. It sits in EU guidance. It is a different instrument from a U.S. parental-consent flow and from a UK data-minimization duty.
UNICEF’s materials cross these regions as child-rights guidance. Version 3 and its checklist call for privacy protections, proportionate collection, and explanations children can understand. A company can follow the UNICEF checklist and still have to meet COPPA, UK children’s-information duties, or EU law, depending on where it operates and where the child is. A company can also meet a local legal minimum and still fall short of UNICEF’s standard on plain language or on how little data it collects. Those are separate judgments, and families should record them separately.
Also popular on Amazon
- Portable and powerful USB-C HUB: BENFEI USB Type-C HUB, with super-soft and knot-free silicone woven design cable, meets most mobile office needs. Compact, lightweight, stylish, and powerful portable USB C Hub equipped with 1 x HDMI port, 1 x 100W charging, and 3 x USB ports. 18-month warranty, 24-hour response, to ensure you feel at ease when using our product.
- Design centered on comfort and reliability: Thanks to BENFEI's end-to-end in-house cable production capability, in-house PCBA and assembly capability, using the industry's most advanced silicone woven design and process, 20cm cable in length, no knots, super-soft, the HUB is easy to use in all scenarios: laptop, tablet, stand etc. Super-soft, 25000+ life cycles, to meet your daily carrying and office needs.
- 100W Charging: Support up to 90W USB C pass-through charging via Type-C port to keep your laptop powered. 10W is reserved for other interface operations. No data and video function on the Type-C port.
- 4K HDMI Display: The HDMI port supports media display at resolutions up to 4K 30Hz, keeping every incredible moment detailed and ultra vivid. Please note that the C port of the Host device needs to support video output.
- Transfer Files in Seconds: Transfer files and from your laptop at speeds up to 10 Gbps with USB A 3.2 port. Extra 2 USB A 2.0 ports are perfectly for your keyboards and mouse.
| Setting | Official source | What it contributes to this question | Boundary |
|---|---|---|---|
| United States | FTC COPPA FAQ | For covered operators, direct parental notice about covered collection, use, or disclosure of children’s personal information, and parental consent | U.S. duties for covered operators and circumstances. The FAQ is not a worldwide chatbot code |
| United Kingdom | ICO guidance on processing that can affect children’s rights | Attention to privacy, informed choices, transparency, and data minimization | UK children’s-information guidance, written for organizations and usable by families as a reading guide |
| European Union | European Commission 2025 guidelines on minors online | AI chatbots and companions may affect minors’ privacy, safety, and security; age assurance is described as one possible way to reduce exposure to risks | EU guidance. It does not state a retention period for any particular chatbot |
| Child-rights guidance across countries | UNICEF Guidance on AI and Children 3.0 and its checklist | Purpose-specific and proportionate collection, lawful use with informed consent, age-appropriate explanation, and protections around relational AI | Guidance for evaluating practice. It does not replace the law that binds a given provider or reader |
For a child living outside the United States, the United Kingdom, and the European Union, none of the three legal sources above is automatically the local rule. The service’s privacy notice may still describe rights it offers in more than one country. UNICEF’s checklist remains a set of questions you can ask of any notice. Enforcement at home is a question for a qualified adviser in that place.
What should parents and educators check in the privacy notice?
You need three current documents for the exact app and account the child will use: the privacy notice, any age or child-safety page the company links from it, and the privacy settings visible after login. A marketing page or an old help-center post is a weak substitute. Policies change, and a setting can differ by plan.
Work through the questions below in order. Each one is something to investigate in that product. The list does not assume every chatbot offers the same switch.
- Account information. List what signup requires: age, a parent’s contact details, payment information for a paid tier, or other identifiers. UNICEF flags age, parent contact, and billing details as examples some systems require. Where a field is required, the notice should say why.
- Stored content. Record whether prompts, uploaded files, voice inputs, and transcripts are stored, and any retention period stated for each. If the period is missing, write “not stated.” Leave the cell blank of guesses.
- Training and improvement. Record whether conversation data may be used to improve the service or train models, and whether a setting limits that use. Copy any sentence about what the setting leaves in place.
- Review and sharing. Record whether staff or service providers can read conversations, and what sharing or disclosure the notice describes. The FTC’s orders put use and sharing among the topics companies were asked to explain to the Commission. The public description available to a family is the one in the notice.
- Age rules. Record the stated minimum age, whether the service says it is meant for children, and what the company says happens if a user is under that age. A published age rule is the company’s condition of use. These sources do not show that every age field stops a younger child from opening an account.
- Deletion and parent tools. Record the deletion path, what it covers, and the exceptions. Then list the parental controls the product actually names, such as viewing history, turning history off, or approving chats. A control that only limits minutes of use leaves training, sharing, and retention untouched.
Educators should run the same six checks on the vendor that processes student prompts. The FTC’s September 2025 orders address companies that provide consumer-facing companion chatbots. A classroom product can still receive what children type, under its own notice and its own operator. Ask the school for the vendor name and that vendor’s notice. If students also use a consumer companion on their own time, keep that second notice separate. One vendor’s age rule does not describe the other product.
After the six checks, set the notice next to UNICEF’s expectations: a specific purpose, collection limited to that purpose, lawful use with informed consent, and an explanation a child can understand. A miss on one of those is a reason to limit use or to choose a different service. A miss on plain language is not, by itself, a completed finding that a particular statute was broken. That legal conclusion depends on the operator, the child, and the rule that covers them.
What information should children avoid sharing with AI?
Children should keep out of the chat any detail they would not want the company behind the bot to receive and handle under that company’s stated rules. The details with the clearest downside, given the categories UNICEF highlights, are health information, feelings tied to identifiable events, facts about other people, and context that identifies the child offline.
A household list can be short enough to put on the fridge:
- Full name, home address, school name, teacher name, and routines such as the street a child walks alone.
- Health conditions, medications, therapy, injuries, or appointments.
- Passwords, codes, payment numbers, or a parent’s login.
- Private facts about another child, including photos or voice recordings of classmates.
- Anything the child would be upset to have repeated to a person reviewing chats or carried into a later version of the product.
For younger children, the same list is easier to teach in the positive direction: use the bot for general questions, and talk with a trusted adult about anything personal. A child can ask how asthma works without naming a clinic. They can ask how to handle an argument without naming the friend. That practice applies UNICEF’s emphasis on minimizing children’s data and on explaining privacy in an age-appropriate way. It is a household rule built from that guidance, not a slogan quoted from UNICEF. It reduces what this conversation contains. It leaves account data, old chats, and training settings where the notice already put them.
UNICEF’s parenting material on approaching AI with children is the family-facing companion to the version 3 policy guidance, and it is worth reading beside the notice rather than instead of it.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Adults should follow the same line. Pasting a child’s medical note, a school report, or a full name into a bot to “summarize it” puts that child’s personal information into a conversation, this time from the adult account. The notice that applies is the notice on the account that did the pasting. That account may allow review or model improvement on terms a child profile would describe differently. Read the account you are about to use before you paste.
Also popular on Amazon
- 【Read before You Buy】 Note: This adapter is USB 2.0 version and cannot transmit HDM, VGA, or other video signals. Nevertheless, they are fully capable of charging devices,duplicating files, and can be seamlessly integrated with USB-C earphones, wireless network adapters, and SD card adapters.
- 【Easy to Use and Safe】Thanks to modern manufacturing techniques, we've made these adapters sufficiently compact so that they can remain inserted in your device without noticeable protrusion or inconvenience. This means you don't have to constantly plug and unplug them, reducing the likelihood of wear and misalignment, and making them more convenient to carry around.
- 【Reliable Quality】Engineered for enhanced durability, durable aluminum alloy shell material,get excellent wear resistance and heat dissipation,the male USB C and female USB connectors & the male USB and female USB C connectors are designed to be plugged and unplugged up to 20000 times. Provides data transfer, device charging, memory space expansion, and additional computer peripherals.
- 【Think more for you】Do you need to connect a USB-C-enabled device to an odd device that only has USB-A ports? Then these USB-A male to USB-C female adapters are made for you.Simply insert the adapter into the USB-A port on the device, and you can instantly connect your USB-C device to the device that originally didn't have the correct connection port.
- 【Trusted xiwxi Quality and Service】As a responsible brand seller, you will receive xiwxi's 12-month Quality assurance and lifetime friendly customer service, you're not just buying a product,you're investing in peace of mind with all-round quality and exceptional customer service.
How can you compare two chatbots on children’s data?
Compare services on the same dimensions, using each company’s current official privacy documentation. The sources gathered here do not support a ranking of chatbot products, and they do not record any company’s present settings. A label such as “most private chatbot” needs those current notices. Without them it is a guess.
Use one row per service and fill it only from the notice, the child-safety page, and the settings screen:
- Information required at signup.
- Whether conversation content is retained, and the period or rule the notice states.
- Whether chats may be used for product improvement or model training, and whether that use can be limited.
- Human review and third-party sharing the policy describes.
- Age rules and any child-specific protections the company states.
- Deletion, privacy, and parental controls, plus any limits the company states.
Keep two columns in your notes: what the company promises, and what has been independently verified. This article does not verify vendor promises. A promise still matters, because it is the commitment you can cite when you use a setting, ask for deletion, or decide the service is too vague for a child. When a toggle and a policy paragraph disagree, write both down and ask the company which one controls the child’s account. These sources do not establish that the friendlier label overrides the broader legal text.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Re-read both notices when the product updates its model or its terms. The September 2025 FTC inquiry shows that companion-chatbot data practices were non-obvious enough for a U.S. regulator to compel answers. It is not a standing score you can cite in place of the policy dated on the site you are about to allow.
What should you do when the privacy notice is vague or incomplete?
When the notice does not say how long chats are kept, whether they are used to train models, who can read them, or how deletion works, the family is missing the facts those choices require. The decision that matches the public record is to pause that service for the child, or to limit it to impersonal questions, until the company answers in the notice or in a direct reply. The September 2025 FTC documents supply questions and an investigative rationale. They do not supply a per-company finding you can use as a substitute for the missing paragraphs.
Several branches come up often:
- No privacy notice in the app. You then lack the provider’s account of collection, retention, sharing, and deletion. UNICEF’s call for an age-appropriate explanation is unmet on the face of the product. Wait until the notice is available.
- A notice written only in dense legal language. You may still extract the six facts above, and you should. You can also conclude that a child cannot understand the explanation. That is a separate fail under the UNICEF standard, even when the underlying collection looks narrow.
- A delete button with no description of leftovers. Use the button if you want the control it actually offers, and record the exceptions the notice names. The cited sources do not show that any one control erases every copy, including backups or material already used to improve a model.
- A clear notice whose terms the household rejects. Purpose, retention, sharing, and training either fit the household or they do not. Where they do not, another service, or no chatbot, is the option the documents leave open.
Where the notice is specific and acceptable, the matching exercise is straightforward. If the stated purpose, retention, sharing, and training terms fit what the household will accept, and if the age rules fit the child, use the settings the notice describes and read them again when the terms change. UNICEF’s checklist — purpose limits, minimal collection, informed and lawful use, and age-appropriate explanation — is the yardstick for that fit. COPPA, the ICO children’s-rights materials, or the European Commission guidelines are the yardstick for legal duties, each in its own jurisdiction.
The practical outcome is specific. A chatbot collects what the child puts into the conversation and what the account requires. It processes that material to answer. The provider’s current rules decide storage, review, sharing, improvement, and deletion. Families who read those rules, teach children to withhold identifying and sensitive detail, and set aside services that will not explain themselves are acting on what the official record actually supports.
Frequently Asked Questions
Was UNICEF’s 2025 AI guidance informed by research with children and caregivers?
UNICEF’s Office of Research – Innocenti says the updated Guidance on AI and Children drew on emergent evidence that includes a twelve-country study with children and caregivers. That figure describes the evidence base for the guidance. It is not a count of how many children use chatbots, and it is not a measure of how often their chats are stored or shared. The report series is introduced on UNICEF’s policy-guidance page: https://www.unicef.org/innocenti/reports/policy-guidance-ai-children
What is a neutral age screen in the FTC’s COPPA FAQ?
The FTC’s COPPA FAQ discusses neutral age screens as their own topic, alongside direct notice to parents and parental consent, for operators covered by the U.S. rule. An age box on a chatbot signup page does not show that the parental notice was given or that consent was obtained. Families who need the Commission’s description should read the neutral-age-screen portion of that FAQ. Adding an age field does not, by itself, place an operator outside the United States under COPPA.
Does a free chatbot avoid regulators’ questions about monetizing engagement?
A price of zero leaves the FTC’s question about monetization of engagement unanswered. Engagement can still be a business issue when a family never enters a card. Read the privacy notice for sharing, product improvement, and any account of how the company monetizes use of the companion. The price on an app listing is a different fact from those sections.
Does EU age assurance erase a minor’s chatbot messages?
Age assurance and deletion solve different problems in the European Commission’s 2025 guidelines on minors online. Those guidelines discuss age checks as a way to reduce who encounters risky services, including AI companions. They do not define age assurance as a feature that wipes transcripts, limits training, or shortens retention once a minor is in the chat. Any deletion path still has to be read in the provider’s own notice, including the exceptions that notice lists.
Free tools Windows power users keep installed
One-click scans. No signup required.
Which country’s rules apply if a child uses a chatbot while traveling?
COPPA duties, the UK ICO’s children’s-rights materials, and the European Commission’s 2025 guidelines each stay attached to their own legal setting when a phone crosses a border. A trip does not merge them into one chatbot rule, and a signup flow run by a foreign company does not close off questions a family may raise under the rules at home. Read the region named in the service’s privacy notice, then use the official guidance for the country where the child lives to frame what you ask the provider. Those public documents are not personal legal advice about one account.
