Is Kindroid safe? Who can decrypt your chats, and when

The Morse teamPublished October 29, 20268 min read

As of October 2026, Kindroid is safe from other users but not private from Kindroid itself. Its privacy policy says your chats are "encrypted at rest and in transit," then adds that "Kindroid is not end-to-end encrypted," because its systems need to read your messages to answer them, and authorized staff may read excerpts its safety systems flag.

We read Kindroid's privacy policy (effective October 22, 2026), its terms of use (effective October 6, 2026), its moderation, age verification, memory, and plans pages, and its update log on October 23, 2026. Kindroid says more than most companion apps do. It names when it decrypts chats, and it says flatly that it hasn't sold data.

Kindroid's privacy settings at a glance

Read on October 23, 2026. Every row links to Kindroid's own page.

Who owns Kindroid?

A Los Angeles company called Beautifully Incorporated. Its privacy policy names "Beautifully Incorporated ("Kindroid")" at 5551 Hollywood Blvd #1277, Los Angeles. The same policy says "We are based in the United States, and we and our Service Providers process Personal Data there." Kindroid runs on iPhone, iPad, Android, and the web, all from one account (Kindroid).

Can Kindroid staff read your chats?

Some of them can, in specific cases. Kindroid's home page says "Our support team can't read your chats" (Kindroid), and that's accurate as far as it goes, but support is one group, and the privacy policy names others (Kindroid):

Kindroid's policy says it decrypts content for human review or disclosure only when its safety systems have flagged it, when you ask (as in a moderation appeal), or when the law or a contract requires it. Its moderation guidelines, updated October 9, 2026, say "no human reads any content" during automated detection, and that a person sees flagged content only if you appeal a lock.

A court order reaches whatever a company stores, and Kindroid stores your chats until you delete them. We cover how that has played out for other AI apps in can AI chats be subpoenaed.

Does Kindroid train its AI on your chats?

Kindroid's privacy policy never says either way. The word "train" doesn't appear in what we read on October 23, 2026. What it does list among its uses is "Improving the Services, including testing, research, internal analytics and product development" (Kindroid). It also says it may create "aggregated, de-identified or anonymized data" from what it collects and use it "to analyze, build and improve the Services."

The terms of use go further for what you write in private. They let Kindroid "de-identify and/or aggregate your Private User Content" and then use the result "for any purpose, without acknowledgement or compensation to you." The licenses you grant are "royalty-free, fully paid, perpetual." Kindroid keeps its rights in "the AI models, algorithms, and systems used to create the AI Generations," and neither document names an outside AI model provider.

Kindroid hasn't ruled training out. For how the big chatbots answer the same question in writing, see which AI companies train on your data.

Does Kindroid moderate private chats?

Yes, by machine. The moderation guidelines say an AI system scans recent chats, calls, and media requests against three red lines: imminent self-harm, imminent harm to others, and child sexual abuse material. Older chats and media aren't scanned. A warning always comes before an account lock, and if violations continue, "the account is locked automatically." A person decides an appeal, and only after you give "explicit consent" for a Trust and Safety team member to see the content.

How long does Kindroid keep your chats?

Until you delete them. The privacy policy keeps "your chats, memories, Kindroid settings and media until you delete them or your account." Its section for UK, EEA, and Swiss users adds that deleting a single message, journal entry, or piece of media takes effect at once, while deleting a Kindroid or your whole account "takes effect 24 hours after you request it." After deletion, Kindroid keeps "a limited set of account and registration records" to stop banned users from coming back, plus what the law requires for tax and accounting.

Stopping payment doesn't delete anything. When the free preview or a subscription ends, the account turns read-only (Kindroid), and the update log says "Your Kins and existing conversations are not deleted when your interactions run out."

Some memories are hard to remove without deleting the whole Kin. Kindroid's memory guide, updated October 13, 2026, says you can edit journal entries and Learned Context, but long-term memories can't be edited; you can only mark one Deprioritize so your Kin leans on it less. To clear cascaded memory, you turn on Reset Cascaded Memory when you confirm a chat break.

To keep a copy before you delete, the privacy policy offers a self-serve export from Profile, under Preferences, once every 30 days.

How does Kindroid check your age?

Kindroid is for adults: its terms say the service is "only for use by individuals who are at least 18 years of age," and the company can ask for proof of age at any time. Its age verification page, updated October 9, 2026, describes layers:

• Paying users are verified through the payment. Time from a gift code doesn't count.
• Free users can be flagged by a system that estimates age from behavior, and then see only the verification screen until they pass.
• Flagged users verify with Yoti, by a selfie or a government ID. The selfie check passes only if Yoti estimates you're 21 or older, and you can start up to 5 attempts in 24 hours.

Kindroid says it "only receives a pass or fail result and does not receive or store any images or identity documents," and that Yoti deletes biometric data after the check. Kindroid offers no parental consent process (privacy policy), so it isn't a choice for anyone under 18.

Is it safe to tell Kindroid sensitive things?

Kindroid's general policy says its services "are not designed to collect or process sensitive personal information" such as health details or sexual orientation, and that sharing them is at your own risk (Kindroid). Its section for UK, EEA, and Swiss users says the same kind of data, if you put it in chats or character settings, is used only to run the service and is "never" used for advertising. The policy lists analytics partners among those it discloses data to, and its mobile apps pass your advertising choices to an attribution provider, without naming either.

So the policy treats a confession to a companion differently depending on where you live. For the same question across other apps, see are AI chats private.

How Morse keeps the decryption key away from your chats

Kindroid can decrypt your chats because it holds both the chats and the keys on its own servers. Morse splits them. Your chat history, the characters you make, and your personas are stored encrypted in your browser, and our servers don't keep messages or log their content (privacy policy). Our privacy policy puts it this way: storage encryption "is not end-to-end encryption against us: we hold the key, but we never receive the data it unlocks."

• No flagged excerpts to pull. A copy of your browser data alone can't be read, because unlocking takes a key that never leaves your device plus one our server hands you at sign-in, and everything locks again when you sign out.
• No de-identified copies to reuse. Morse doesn't train AI on your content, sell your data, or show ads, and runs no advertising trackers or third-party analytics.
• Little to hand over. Your account holds sign-in details, your plan, your balance, and a usage ledger with no content, so we usually have very little to give authorities.

When you want a conversation that even the AI host can't open outside a verified machine, start it in Chat:

• E2EE, for anything you'd never want decrypted by a company. Your browser checks the enclave's hardware attestation before anything leaves your device, seals the message for that enclave alone, and decrypts the reply itself; our servers relay ciphertext they can't read. The enclaves are run by third-party providers, not by Morse, and the guarantee rests on the enclave hardware and attestation working as designed. E2EE comes with Pro, $19 a month.
• Private, the default, when you want no host keeping a copy. Open-weight models like DeepSeek, Kimi, and GLM run under Morse's zero data retention agreements with its inference, search, and other providers.
• Anonymous, for GPT, Claude, or Gemini without the provider knowing it's you. We strip your account, email, and IP address, though the provider may keep a message for a limited time to check for abuse, under its own policy.

Because your history is stored only in your browser, clearing it deletes your chats for good, so keep a copy of anything you want. How the enclave check works is in our guide to end-to-end encrypted AI chat.

Send a message from the Morse home page without an account, or start free and open Characters with no card.