Is Chai AI safe? Its age rules, content filters, privacy and the laws that may apply

The short answer

Chai's Safety Center and license agreement say the app is for users 18 and older, and the Safety Center says Chai uses a third-party provider, Persona, for age checks by selfie or photo ID. It also says Chai's models and real-time safety systems are designed to refuse sexually explicit content, self-harm content and content that sexualizes minors.

The Safety Center describes a self-harm protocol that scans messages and shows a crisis referral notice. Chai's privacy policy says chat conversations are used to fine-tune its AI models after personal identifiers are removed, and the policy does not describe a training opt-out. Chai AI was among 13 companies that groups of state attorneys general wrote to in August and December 2025; those were warning letters, not lawsuits.

Who runs Chai

As of October 22, 2026, Chai's privacy policy (last updated October 12, 2026) names Chai Research Corp., a Delaware corporation with an address in Palo Alto, California, as the data controller.

Its End User License Agreement (last updated July 28, 2026) names the same company as licensor and says it is governed by California law. The App Store and Google Play listings name Chai Research Corp. as the developer, and Google Play gives the same Palo Alto address.

Minimum age and how it is checked

Chai's Safety Center says users under 18 are not permitted, and the license agreement says the app is strictly for users 18 and older. Under the agreement, users confirm they are adults and that they gave accurate age verification.

The Safety Center says Chai uses Persona, a third-party identity provider, for age assurance, with a choice of a selfie age estimate or a government photo ID check. It does not say which users must complete the check or when. It also says a paid subscription is required in some regions, without naming them.

The Safety Center says the app is rated 18+ on both app stores. As of October 22, 2026, the App Store listing shows an 18+ rating, with frequent mature or suggestive themes and infrequent sexual content or nudity among its descriptors. The Google Play listing shows a Mature 17+ rating with sexual themes among its descriptors.

Chai's Safety Center, license agreement and privacy policy do not describe a teen mode or parental controls.

Content rules and filters

The Safety Center says Chai's models and real-time safety systems are designed to refuse sexually explicit content, self-harm content and content that sexualizes minors. It says users can filter mature content in search, hide explicit messages and block specific characters.

The license agreement allows adult roleplay with mature themes and fictional violence, as long as it stays within its prohibited-content rules. As of October 22, 2026, those rules bar:

• Sexually explicit images or videos.
• Child sexual exploitation material, including simulated or synthetic depictions.
• Content that promotes suicide, self-injury or severe eating disorders.
• Non-consensual intimate material, including deepfake pornography.
• Terrorist and extremist material, gratuitous or highly realistic violence, and instructions for making dangerous weapons.

Reporting and moderation

The Safety Center says users can report any AI character or message in the app, that reports are anonymous, and that a character's creator is not told who reported it. The license agreement says reports are reviewed by a Trust and Safety team.

The agreement says Chai may prescreen or delete user contributions at any time without notice, while also saying it has no obligation to monitor them. It says breaking its content rules can lead to immediate account termination, preservation of data and possible reports to law enforcement.

Crisis and self-harm features

As of October 22, 2026, Chai's Safety Center says:

• Each user message is scanned in real time for self-harm content.
• When such content is flagged, the system refuses to respond or replaces the reply with a safer one.
• When a user expresses intent to self-harm, the app shows a notice referring them to crisis service providers.
• The Safety Center links to Find a Helpline, a directory of crisis lines in more than 175 countries, and lists emergency numbers including 911 in the U.S.

Payments and refunds

As of October 22, 2026, the App Store lists in-app purchases including bundles of Pieces, monthly and annual Premium subscriptions and weekly and monthly Ultra subscriptions, priced from $2.79 to $134.99. Both app store listings say the app contains ads, and the privacy policy says it uses personal data to show ads to users on the free plan, based on their consent.

The license agreement does not set out subscription or virtual-currency terms. The refund provision we found covers refunds of the purchase price through the app store if the app fails to meet its warranty.

Privacy: what Chai collects

As of October 22, 2026, Chai's privacy policy lists:

• Email and basic details from a linked Google, Apple or Facebook account.
• Chat messages, which the policy notes may include sensitive information users choose to share, and the chatbots users create.
• Emails to support and the data needed to manage paid subscriptions.
• IP address, access times, browser, operating system, internet provider, essential cookies and activity logs tied to the user ID.

Privacy: training, staff access and sharing

The privacy policy says Chai uses data from in-app conversations to improve and fine-tune the AI models behind its chatbots, and that personal identifiers are removed first. It says messages may be used to improve the chatbot for all users. The policy does not describe a training opt-out; it lists a right to object to processing for users in the EU and UK.

The policy says personal data is accessible only to specifically authorized staff and to suppliers acting as processors. It does not say directly whether staff read chats.

It says data is shared with Chai's suppliers, which it does not name, and with authorities where the law requires. For users who installed the app through an ad, it says usage patterns are shared with advertising partners to measure campaigns. It says data is processed and stored in the United States. We did not find a statement on whether Chai sells personal data.

The app store privacy labels are Chai's own statements. The App Store label lists usage data as used to track users, and purchase history and name as linked to the user. Google Play's data safety section says the app shares location, personal info and other data types with third parties, and that data is encrypted in transit.

Deleting your account

The privacy policy says users can delete their account in the app under Profile, then Settings, then Delete Account, which permanently deletes the account and associated personal data, subject to its retention rules. It lists hello@chai-research.com for privacy requests and dpo@chai-research.com for its data protection officer.

The policy says data is kept no longer than needed, and in any case not beyond the life of the account plus five years after deletion, with longer retention possible for legal claims or legal obligations.

Government actions

Chai was not among the seven companies that received the FTC's companion chatbot orders in September 2025. We did not find an FTC or state attorney general lawsuit naming Chai Research in official sources as of October 22, 2026.

Chai AI is named as a recipient in two multistate warning letters from state attorneys general. Neither is a lawsuit:

• August 25, 2025: 44 attorneys general, in a coalition announced by Tennessee's attorney general, wrote to 13 companies, including Chai AI, asking them to stop sexually inappropriate conversations with children and to adopt policies against sexualizing children.
• December 10, 2025: 42 attorneys general, led by Pennsylvania, wrote to 13 companies, including Chai AI, asking for more safeguards, safety testing, recall procedures, clear warnings to consumers and a commitment to changes by January 16, 2026.

Which chatbot laws may be relevant

No law names Chai. Whether a particular service counts as a companion chatbot under a given law is a legal question. These state laws set duties for companion chatbot operators:

• California SB 243, in effect since January 1, 2026: operators must give a clear notice that the chatbot is AI when a reasonable person would be misled, keep a protocol to prevent suicide and self-harm content that refers users to crisis services, publish details of that protocol on their website, and tell users companion chatbots may not suit some minors. Injured users can sue.
• New York's AI companion law, in effect since November 5, 2025: operators need a protocol that makes reasonable efforts to detect suicidal thoughts or self-harm and refer users to crisis services, and must tell users they are not talking to a human at the start of an interaction and at least every three hours.
• Washington HB 2225 and Oregon SB 1546, both effective January 1, 2027: operators must disclose that the chatbot is AI (in Oregon, when a reasonable person could think it is human), run and publish a suicide and self-harm protocol with crisis referrals, and add protections for users the operator knows are minors (in Washington, also where the chatbot is directed to minors; in Oregon, also where the operator has reason to believe a user is a minor).

We are not saying Chai is covered by, or out of compliance with, any of these laws. It may have practices it has not published, and whether a law applies to it is a legal question this guide does not decide.