Character.AI Safety Settings

Character.AI safety settings help parents understand access, privacy and the limits of this service. Character.AI is an entertainment platform for simulated characters and creative stories. A character may resemble a fictional favourite or use an invented personality. Young people may enjoy the imaginative interaction. However, the current experience for under 18s differs substantially from older online tutorials.

Start with the age warning below. Then review the relevant controls and practical family advice. This guide explains protections and limits rather than recommending that children join a platform.

Age Rating and Child Safety Warning

Character.AI’s current help describes Reading Mode for users verified as under 18. They may browse characters and view platform content, but chat features are unavailable. Older daily chat allowances should not be treated as the current setup. Family Connect offers optional screen time sharing, as explained below, but does not restore restricted chat access. Read the current provider guidance.

The account terms set a minimum age of 13, rising to 16 for citizens or residents of the EEA or UK. This includes Ireland. However, meeting the account age does not grant permission to use restricted chat features. Parental consent does not waive these minimum ages. Read the Character.AI account terms.

Character.AI first uses account and activity signals for age assurance. An adult correcting an under 18 result can use Persona selfie verification, with ID requested if needed. Character.AI says Persona handles the submissions and deletes biometric or ID evidence within seven days. Read the actual notice before supplying evidence. Use only the provider’s official route and never a link sent by a stranger. A successful age check does not establish emotional readiness or suitable content.

Character.AI Access Restrictions

Character.AI: Connect a Parent for Screen Time Insights

Review Character.AI safety settings alongside Family Connect, which shares screen time. It cannot show messages, viewing content or enforce time limits. Also, either person can disconnect.

Adult verification for parent linking uses k-ID. This is separate from Persona verification used to correct an account’s age assessment. Therefore, read the notice for the actual process. A pending parent invitation expires after 14 days. For verification help, use the provider’s linking support guidance.

Teen setup

• Teen: Open Settings, then Parental insights.
• Enter the trusted adult’s email and send the invite. Wait for the connection to show Linked parent.

Parent setup

• Parent: Open the genuine invitation email and follow its linking instructions. Check the destination before supplying identity evidence.
• Complete the adult verification and open Family Connect to review screen time and the weekly digest.
• If linking stays pending, reopen the teen’s profile and check again. Ask official support if it still fails.

Why it matters: The parent can see accumulated screen time, not just activity after linking. However, the teen can unlink and the parent receives a notification. Keep device limits and regular conversations in place.

Set Device Limits with Family Link

For an eligible supervised Android device, a parent can limit access to the app. This is a device control rather than an AI content filter.

• Parent: Open Family Link and select your child.
• Open Screen time, Time limits and App limits. Select Character.AI.
• Choose Set limit and save the agreed allowance, or use Block when access is unsuitable.

Why it matters: A device limit cannot make an ineligible account acceptable. Also review browser access and other devices, because an app limit does not cover every route.

Use Screen Time on iPhone or iPad

Family Sharing allows parents to manage a child’s Screen Time. Set this up with the child’s own Apple Account.

• Parent: Open iPhone Settings, then Screen Time and select the child under Family.
• Open App Limits, select Add Limit and select Character.AI if installed. Set the allowance and add the limit.
• Review Downtime and protect changes with a Screen Time passcode the child does not know.

Why it matters: Review exceptions and other ways to reach the service. However, time limits do not inspect generated answers or replace the provider’s age requirements.

Character.AI Safety Settings

Some Character.AI safety settings vary by account, device or country. A setting on the website may differ from the app. Therefore, check the confirmation and its scope before relying on any change. In addition, keep access restrictions separate from content, privacy and billing choices.

Character.AI: Use the Genuine Age and Reading Mode

Age assurance aims to give the account its age appropriate experience.

• Keep the account’s date of birth accurate. Do not use an adult’s identity for a young person.
• Next, an eligible adult correcting an error can open Settings, Advanced and Verify Age in the app or website. Read the verification notice before proceeding.

Why it matters: A verification result does not certify content as safe. Children should not submit identity documents through unsolicited links.

Character.AI: Turn Off Model Training Where Available

The documented choice applies in the EEA, including Ireland, and the UK.

• In the app, open Settings and Data & Privacy. Deselect Improve the Model for Everyone.
• Next, on the website, open Settings, Account, Manage Account & Data and Model Improvement. Deselect Improve the Model for Everyone.

Why it matters: The choice covers new chat content for generative model training. Feedback and other service improvement can remain separate. US availability should not be assumed.

Character.AI: Review Muted Words

Muted words is a conversation preference for words someone would rather avoid. However, it is not a locked parental filter and cannot prevent every harmful topic.

• Open Profile, select Settings and choose the Muted words tab where available.
• Next, add the words you want to avoid. The help article’s updated limits are four words for free accounts and twenty for c.ai+. Check the limit displayed on the actual account.
• Finally, if the tab is missing, ask the official help centre about availability rather than relying on an older tutorial.

Why it matters: Complex or multilingual words may still appear. Therefore, stop a distressing conversation even when a word preference is active. This control does not restore chat access for users restricted to Reading Mode.

Character.AI: Block Another Creator

Blocking hides the person’s content in both directions.

• Open the person’s profile, select the three dot menu and Block.
• Next, to review the list, open your own profile, the gear icon for Settings, then Blocked.

Why it matters: Blocking does not submit a report. Existing conversations can remain readable even though new messages to those characters are unavailable.

Character.AI: Keep Character Creations Private

A character’s audience is separate from the account’s chat history.

• When creating or editing a character, review its visibility before saving. Choose Private for access limited to your account.
• Next, treat Unlisted as access for anyone with its link. Avoid putting private details into descriptions, greetings, definitions or shared images.

Why it matters: Private does not mean the provider cannot process the content. Also review scene, voice and feed audiences separately.

Character.AI: Download Account Data

A data export can help an account holder review information before closing an account. This is separate from deleting it.

• On the web, open Profile, then Profile Settings and Account.
• Select Manage Account & Data, then Export Data. Read the confirmation.
• In the mobile app, open Profile, tap the Settings icon, then Legal & Privacy and Export Data.
• Check the account email for the download link. Allow three to four weeks for processing and download within seven days of the email.

Why it matters: An export contains private information. Store it securely, share only what is necessary and do not rely on it for an urgent safety report.

Character.AI: Remove Pinned Personal Details

Pinned Memories keep selected messages available to a character. Review them if a conversation includes identifying or sensitive details.

• On the web, open the affected chat and its Chat Details panel at the top right.
• Select the pinned message you want to unpin and use its Unpin control.
• In the mobile app, open the chat’s three dots menu to review pinned memories. Long press a pinned message to unpin it.

Why it matters: Unpinning changes remembered context. It does not delete the original message, other memory features or records retained by the provider. Also review personas and any saved Lorebooks.

Character.AI: Report Harmful Content

Report a harmful item separately from blocking its creator. A report asks the platform to review a possible rule violation.

• Open the affected user, character, chat or post and select its flag function.
• Choose the relevant reason and submit the report with enough context for review.
• If the flag is unavailable, open the official Character.AI help centre and select Submit a ticket.

Why it matters: A report is not an emergency response or a guarantee of removal. Keep the item link and confirmation, avoid forwarding harmful material, and seek immediate human help where needed.

Character.AI: Manage Website Subscriptions

For a website purchase of c.ai+, the provider documents a customer portal. App store purchases follow the separate store instructions below.

• Open Settings, then Account and Manage. Check whether the plan says an app store manages it.
• For a website plan with a Manage button, open the customer portal and enter the email used for the account.
• Open the secure customer portal link in that email, manage or cancel the renewal, and retain the confirmation. Check spam if the email is missing.
• For a website c.ai lite plan or a missing control, use Submit a ticket in the official help centre and choose Payments and purchases.

Why it matters: Deleting the app and cancelling a subscription are separate actions. Check the actual billing provider and renewal date. Buying a plan does not add parental supervision or establish child suitability.

Character.AI: Review the Chat Memory Box

When reviewing Character.AI safety settings, check the chat’s Memory text box for saved personal information. Review it separately from pinned messages.

• In an existing eligible chat, select the Character avatar at the upper right and open Memory.
• Next, review the free form text box. Remove addresses, school details, health information or other identifying text you previously entered.
• Follow any displayed save prompt, then reopen Memory to check the text. If your version differs, ask official support before sharing more private information.

Why it matters: This edits the text box, not every memory system or retained record. In addition, review pinned messages, personas and Lorebooks separately. Do not assume a changed memory erases the original conversation.

Character.AI: Review Lorebook Privacy and Removal

Lorebooks store information characters reuse. Creation and editing are c.ai+ beta features, so availability may change.

• Open Profile, then the Lorebook tab and the saved Lorebook.
• Select Lorebook Settings, then Visibility. Choose Private when only you should access it. Private Lorebooks require c.ai+ and work with private Characters.
• For a public Lorebook, turn off Let others link their Characters if unwanted.
• To remove an entry, open its three dots, choose Delete and confirm.
• To delete the whole Lorebook, return to the profile’s Lorebook tab. On the web, use its three dots and Delete. In the app, long press it and choose Delete.

Why it matters: For public entries, Hint Only hides the body but exposes titles and keywords. It does not make the Lorebook private. Also, audience changes cannot recall screenshots or copies.

Character.AI: Review Saved Comics Before Sharing

Comics can use recent chats and the active persona to create a story. As a result, a creation may reveal details from an apparently private interaction.

• In the mobile app, open Chats, then Gallery and the Comic tab.
• Open a saved comic and check its images, captions and dialogue for identifying details.
• Next, use the page’s three dots and Delete this page to remove an unwanted page, or Report for harmful content.
• For the whole comic, use its three dots menu in the gallery to review the deletion, download or reporting choices. Avoid Publish when it should remain unshared.

Why it matters: Publishing puts the comic on the community feed and profile. Therefore, review every page first. Also check the displayed Charms cost before generating or regenerating content.

Character.AI: Delete an Unwanted Account

Deletion is permanent, so consider necessary evidence and billing first.

• On the web, open Profile, Profile Settings, Account and Manage Account & Data.
• Next, select Delete Account and read the confirmation. For app differences or lost access, use the official help centre.

Why it matters: Do not rely on uninstalling to close the account. A privacy request and subscription cancellation can require separate action.

Character.AI: Review Camera, Microphone and Location Permissions

Device permissions control access to sensors and selected files. For example, give only the access needed for the current task.

• On Android, open device Settings, Apps and Character.AI. Open Permissions.
• Next, select each permission and choose Don’t allow for unnecessary access. For a needed feature, choose a limited option where the phone offers it.
• Finally, on iPhone or iPad, open Settings and Privacy & Security. Review Microphone, Camera, Photos and Location Services for this app.

Why it matters: Permissions do not remove files or recordings already uploaded. Therefore, keep other people out of voice and camera sessions and review access after updates.

Character.AI: Reduce Notifications and Protect Sleep

Device notification choices can reduce prompts to return. This is a device preference rather than a platform content filter.

• On iPhone or iPad, open Settings, Notifications and Character.AI. Turn Allow Notifications off if unwanted.
• Next, on Android, open Settings, Notifications and App notifications. Select the app and turn off unwanted notifications.

Why it matters: Also review browser notification permissions. However, silencing an alert does not prevent someone opening the service voluntarily.

Character.AI: Protect Sign In and Recovery Access

Use a separate eligible account and keep its recovery method private. However, this is practical account protection, rather than a claim that this app offers a particular security switch.

• Use a unique password wherever you create a password for Character.AI. Also protect the email inbox used for recovery.
• Next, if you sign in with Google or Apple, open that provider’s account security settings and enable its additional sign in verification.
• Finally, sign out on shared computers. Never give another person a login code, even if a message claims to come from support.

Why it matters: For example, a person with account access may read saved conversations. In addition, a secure login cannot stop a user voluntarily sharing private material.

Character.AI: Check App Store Renewals

If the receipt shows an app store subscription, cancel through that store. Deleting the app is a separate action.

• For Apple billing, open Settings, your name and Subscriptions. Select Character.AI and Cancel Subscription if available.
• Next, for Google Play billing, open Play Store, the profile icon, Payments & subscriptions and Subscriptions. Select the subscription and Cancel subscription.
• Finally, follow the confirmation prompts and check the expiry date. For a different billing provider, use the route on the receipt.

Why it matters: Removing an account may not stop a store renewal. Therefore, retain the cancellation confirmation and check the account that actually paid.

Character.AI Safety Settings for Voice Reporting

A voice can imitate a real person, so review its source and audience.

• Open the affected voice in voice selection or search. Expand its detail card and select the Flag icon at the top right to report it.
• Next, when creating a voice, choose Private if it should be available only to you. A public voice can be used by others.
• Finally, for a safety issue without a clear reporting control, open the official help centre’s Submit a ticket link and explain the concern.

Why it matters: A block and a report perform different jobs. Also, private voice visibility cannot remove recordings already shared elsewhere.

Understand What These Controls Cannot Do

A filter, paid plan or character label cannot guarantee a suitable reply. Also, removing a chat may leave separate memories, published creations or provider records. If a needed control is unclear, stop sharing private information and ask the official support service what it covers. Do not infer a parent dashboard or locked child setting from an ordinary account preference.

How to Use Character.AI Safely

If your teen loses access to an emotionally important character, acknowledge the disappointment without describing the bot as a real partner. Make time for human support and a creative alternative. Do not lend an adult account or help change a birth date to restore restricted chat access.

Review Character.AI safety settings with the eligible account holder. Keep addresses, school details, passwords and private photographs out of prompts. In addition, do not upload another person’s image or voice to impersonate them. However, an apparently private conversation still involves a company processing information.

Character.AI Safety Advice and Current Concerns

Keep Human Support at the Centre

For example, a character can mirror the user’s language and produce reassuring replies. However, it does not have a human duty of care. It cannot reliably assess a child’s safety, diagnose a condition or replace a qualified professional. Agree who the young person can contact when worried, including when a bot appears supportive.

Notice Changes in Everyday Life

Ask whether chats interfere with sleep, school, friendships or activities. In addition, look for distress when access stops, pressure to buy more interaction or a belief that only the character understands. These signs invite a calm conversation rather than proving a diagnosis. Therefore, make a practical plan together and seek appropriate professional help when concerns persist.

Respond to Sexual Content and Pressure

If a conversation turns sexual, threatening or coercive, stop it. Next, reassure the child that asking for help is the right response. If a real person pressures them through linked communities or messages, treat that contact separately from the bot. Avoid forwarding intimate images and get safeguarding help for threats or exploitation.

Check Identity, Accuracy and Secrecy

An imitation of a celebrity or fictional hero is not an authenticated person. Similarly, a convincing answer may invent facts or encourage a poor choice. Challenge messages that ask someone to hide the conversation from trusted adults. Use reliable human help for health, safety, money or relationship decisions.

Character.AI Support and Regional Differences

Use the official Character.AI privacy or support guidance for account and personal data concerns. Include the relevant content reference and device details, while leaving out unrelated private information. Ask for a request reference so you can follow up. A report is not emergency monitoring.

For families in Ireland, check the Data Protection Commission’s parent guidance. UK readers can use ICO public guidance, while US readers can consult FTC children’s privacy information. Account eligibility, privacy rights and app classifications are different questions.

For a concern involving a child’s immediate safety, contact the appropriate local emergency or safeguarding service. For other worries, speak with a trusted adult, school safeguarding contact or qualified health professional. CODA provides education and training rather than crisis monitoring. Explore CODA’s support resources for further routes.

Character.AI Safety Settings Checklist

• Confirm the genuine service, account age and regional rules.
• Check app access, browser access and shared devices.
• Review Character.AI safety settings alongside privacy, published creations and saved context.
• Protect sleep, real relationships and a trusted human support route.
• Check charges and cancel unwanted renewals through the receipt’s provider.
• Practise stopping, reporting and asking for help without shame.

Character.AI Safety Settings FAQ

Does a friendly character make this suitable for a child?

No. A familiar image says little about what a conversation will generate. Start with the age requirement and then consider the young person’s needs. For adult services, prevent access rather than using an adult account together.

Does deleting the app delete conversations?

No. For example, removing an app affects the device. Account closure, stored chats, remembered details, published creations and billing can need separate actions. Therefore, read each confirmation and ask the provider about anything it leaves unclear.

Should a parent read every conversation?

For parents, begin with honest, proportionate involvement and agreed boundaries. First, ask about the experience and look at worrying interactions together where appropriate. In addition, explain any safety related monitoring. Above all, a young person should know they can seek help without automatically losing every form of digital contact.

Related CODA Guides

Explore AI Character Chat, AI Safety Settings and Resources for Parents.

References

Age requirements and platform rules: official guidance for the named control.

Character.AI account eligibility: minimum ages and regional account conditions.

Character.AI age assurance data handling: identifies Persona and its stated retention period.

Family Connect setup and limits: optional parent linking, screen time visibility and disconnection.

Parent linking verification help: identifies k-ID for the parent connection process.

Set Device Limits with Family Link: official guidance for the named control.

Use Screen Time on iPhone or iPad: official guidance for the named control.

Use the Genuine Age and Reading Mode: official guidance for the named control.

Turn Off Model Training Where Available: official guidance for the named control.

Review muted word preferences: navigation, free and paid limits, and filtering limitations.

Block Another Creator: official guidance for the named control.

Keep Character Creations Private: official guidance for the named control.

Download account data: current web and mobile export routes and the download window.

Review pinned memories: how to review and unpin messages.

Additional Account and Sharing Guidance

Report harmful content: the platform flag and support ticket routes. Its regulator advice is specific to Australia.

Manage website subscriptions: website customer portal and separate app store billing.

Get help with lite billing: the Payments and purchases support option for website plans.

Chat memory text box: documents the character avatar and Memory route.

Lorebook privacy and removal: visibility, linking and deletion controls.

Saved comic review: gallery navigation, reporting, removal and publication audiences.

Delete an Unwanted Account: official guidance for the named control.

Review Camera, Microphone and Location Permissions: official guidance for the named control.

Reduce Notifications and Protect Sleep: official guidance for the named control.

Check App Store Renewals: official guidance for the named control.

Character.AI voice privacy and reporting: documents voice audiences and the reporting flag.

Privacy and account information: official guidance for the named control.

eSafety advice on chatbot relationships: wider guidance for parents about generated conversations and human support.

About Children of the Digital Age

Children of the Digital Age provides online safety education and training for families, schools and organisations. Our work connects practical guidance with an understanding of online behaviour and cyberpsychology.

Send a message

Contact us about Character.AI safety settings education, parent sessions or school presentations. Tell us which service your family encounters and what you would like to understand.

Last updated: 16 October 2026