Quick Answer
Character.AI memory is controllable but not perfect, and c.ai+ does not remove platform safety rules. As of May 2026, Story Memory, Pin to Memory, Facts, message history, and background memory management can shape a chat. Free users receive core memory features; c.ai+ receives more memory, more pins, and fuller memory visualization.
Character.AI also removed open-ended AI chat for under-18 users beginning in November 2025 and moved them to a separate creative experience. The current Community Guidelines still prohibit pornographic, exploitative, harmful, impersonating, infringing, and other disallowed content across account types.
Source review: August 3, 2026. Memory rollout, account eligibility, moderation behavior, privacy choices, and subscription features can change; verify the signed-in product and current official policy pages.
For the paid-plan decision, see the c.ai+ pricing guide. For a product switch rather than a settings question, compare Character.AI alternatives.
Decision Table
| Question | Current boundary | What to do |
|---|---|---|
| Will a Character remember a detail? | Story Memory and pins raise the chance; recall is not guaranteed | Save the fact explicitly and verify it in the same chat |
| Does a paid plan remove filters? | No; platform policies still apply | Rewrite within policy or report a false positive |
| Can an under-18 account use open-ended chat? | Character.AI announced removal beginning November 2025 | Use the age-appropriate creative experience and current help guidance |
| Is a chat private from Character.AI? | The service processes chat and account information under its privacy policy | Avoid secrets and review regional privacy choices |
| Can generated advice be trusted? | The Terms warn that generated content can be inaccurate or offensive | Verify important claims with authoritative sources and people |
How Character.AI Memory Works Now
Character.AI’s May 21, 2026 memory release describes four related layers:
- Story Memory stores backstory, key events, and other context a user wants carried through the chat.
- Pin to Memory preserves an exact message selected by the user.
- Facts hold details the system has identified or the user has supplied.
- Memory Usage shows how Facts, Story Memories, and message history consume the available memory.
The service also says it compresses or tidies older context in the background as a chat grows. User-written Story Memory and pinned items are protected from that automatic cleanup, but protected storage still does not prove the Character will apply a fact correctly in every answer.
Story Memory and Pin to Memory are available to all users. c.ai+ provides twice as many pins, more memory, and more detailed memory management. The current release ties the improved memory system to PipSqueak 2 for free users and DeepSqueak for c.ai+.
A reliable continuity workflow
- Put only durable facts in Story Memory; do not turn it into a full transcript.
- Pin the few exact messages that change the story or relationship.
- Inspect Memory Usage before a long chat becomes crowded.
- Ask the Character to restate critical constraints before the next scene.
- Correct contradictions in the chat rather than assuming saved memory guarantees compliance.
Do not store passwords, private health details, financial information, or other secrets merely to improve continuity.
What the Content Filter Does
Character.AI’s current Community Guidelines say its systems use automated moderation and human review, while model filters and limits seek to prevent inappropriate outputs. The rules prohibit categories including child exploitation, pornographic content and nudity, real-world violence and gore, hateful or harassing content, promotion of self-harm, impersonation, privacy violations, infringement, illegal instructions, and attempts to bypass safety systems.
A filter can affect a user input, a generated response, a Character’s discovery, or an account. The company can remove content, warn or suspend a user, or permanently ban an account for serious violations.
If a safe roleplay is interrupted
- Remove ambiguous requests that resemble prohibited real-world instructions.
- Make fictional framing, consent, adult characters, and non-graphic boundaries explicit where relevant.
- Split a complex scene into smaller prompts.
- Use the in-product report or feedback path for an apparent false positive.
- Do not use jailbreaks or filter-bypass instructions; the guidelines prohibit attempts to defeat safety systems.
c.ai+ can change chat style, memory capacity, or rollout access. It is not a filter-removal product.
Under-18 Access and Age Assurance
Character.AI announced on October 29, 2025 that it would remove open-ended chat for users under 18, with the transition beginning in November. The company retained an under-18 experience centered on non-chat creative features such as Feed, Imagine, AvatarFX, Streams, and Character or scene creation, subject to current availability.
The age-assurance process evaluates account and usage signals. If it identifies an adult as under 18, the user can follow the offered verification path. Character.AI says Persona generally starts with a selfie and requests an ID only when the selfie result is inconclusive; Persona, not Character.AI, collects that ID under the described flow.
Age, country, rollout stage, and account history can affect what a user sees. Use the official help flow for an incorrect age result rather than opening a replacement account or supplying documents through unofficial channels.
Privacy and Data Boundaries
Character.AI’s privacy policy took effect July 1, 2026. It says the company collects content and usage information to provide, customize, secure, analyze, and improve the service, including training AI and machine-learning models. The policy also describes disclosure to vendors, moderators, analytics and advertising providers, legal authorities when required, and other named recipients.
The regional disclosures state that applicable users can object to or opt out of some model-training use and describe safeguards such as disassociating submitted content from account information and removing directly identifying information before certain improvement processing.
Practical boundary:
- Assume the provider can process chat content; do not treat a conversation as a private vault.
- Review the privacy controls available to the exact region and account.
- Keep personal data out of public Character definitions.
- Remember that Character.AI may preserve the characteristics of a popular public Character after an account deletion to avoid breaking other users’ experience.
- Export or delete information through official account and privacy-request paths where available.
Safety Checklist
Before starting a long or sensitive chat:
- Confirm the account is in the correct age experience.
- Review Story Memory and remove information the chat does not need.
- Treat the Character as entertainment, not a therapist, doctor, lawyer, financial adviser, or factual authority.
- Verify high-impact claims outside the platform.
- Report harmful content and seek a trusted person or emergency service when there is an immediate real-world risk.
For a broader route map, use the C.AI and Character.AI guide. For an alternative creative-writing workflow, compare Character.AI with Claude.
Frequently Asked Questions
Does Character.AI remember previous messages?
Character.AI now provides Story Memory, pinned messages, Facts, and background memory management. These features improve continuity but do not guarantee perfect recall or exact use of every saved detail. Story Memory and Pin to Memory are available to all users; c.ai+ receives larger memory and more controls.
Can users under 18 have open-ended Character.AI chats?
Character.AI announced that it was removing open-ended AI chat for under-18 users starting in November 2025. The company kept a separate under-18 creative experience for non-chat features and uses age assurance to determine the applicable experience.
Does c.ai+ remove Character.AI content filters?
No. Character.AI’s Terms and Community Guidelines apply across plans. The company says automated moderation, human review, model limits, and content filters enforce platform boundaries; a paid subscription is not permission to bypass them.
Are Character.AI chats used to improve its models?
Character.AI’s privacy policy says it may use information to analyze and improve the service, including training AI and machine-learning models. Regional disclosures describe safeguards and opt-out rights in applicable jurisdictions. Do not submit secrets or sensitive personal data to a chat.