Is Tolan safe? What Portola keeps, trains on, and lets parents read

The Morse teamPublished November 12, 20269 min read

Tolan is built to be safe to talk to, but it isn't private from Portola, the company that makes it, as of November 12, 2026. Portola's help center says your conversation history "doesn't expire on its own" and stays until you delete your account (Tolan). The same page answers "Is my data used to train AI?" with "Yes" (Tolan). Tolan's rules ban romantic and sexual roleplay (Tolan). When a chat turns to self-harm, Tolan points you to the 988 Suicide & Crisis Lifeline "instead of engaging" (Tolan).

We read Tolan's privacy policy and U.S. State Privacy Notice (both updated November 6, 2026), its terms of service (June 30, 2026), its community guidelines, its FAQ, its help center, and Apple's App Store listing on November 12, 2026. Portola deserves credit for plain answers. Its privacy FAQ says "Yes" to training in one word, and says the full policy wins wherever the two differ.

Tolan safety at a glance

Does Tolan use your conversations to train AI?

Yes, by Portola's own account. The privacy policy lists, as a purpose for what it collects, "to improve, train, and develop the artificial intelligence, machine learning, and models that we use" (policy). The help center puts it more plainly: "We use what you share and how you use Tolan… to improve, train, and develop the AI models that power Tolan" (Tolan). The terms give Portola a "worldwide, non-exclusive, perpetual, irrevocable" license to store what you send "for the purpose of providing and developing the Service, which includes improving the AI model(s)" (terms).

Neither the policy nor the help center describes a switch to turn training off. Gmail and Google Calendar data are the exception, since the policy says Portola doesn't use them "to create, train, or improve generalized artificial intelligence or machine-learning models" (policy).

Tolan's own website FAQ gives a different answer. It says conversation logs are "only ever used to improve your app experience" and are "never sold or shared without your consent" (Tolan). The privacy policy, read the same day, discloses information to AI service providers, to PostHog for analytics, and to advertising partners (policy).

Is it safe to tell Tolan personal things?

Tolan's terms say not to. Section 8.6 reads, "You must not submit Input that (i) contains personal data or sensitive health information" (terms). Portola's App Store description invites you to talk about "Your feelings: relationships, challenges, stress" (App Store). Portola's U.S. State Privacy Notice says that if you share health information or details of your sex life in conversations, "we will receive that information" (Tolan).

The terms also say Tolan is "for informational and/or recreational purposes only" and that Portola isn't responsible for "feelings of social isolation, loneliness, or addiction to technology" (terms).

Who can see your Tolan chats?

Other users can't, and several kinds of companies and people can. The help center says your chats "are not public, and other users can't browse them" (Tolan). It also says "the AI providers that power your Tolan's responses" process them (Tolan). The policy doesn't name those providers (policy). Tolan's home page says it builds "on top of the leading AI models" and links to an OpenAI case study titled "How Tolan Builds Voice-First AI with GPT-5.1" (Tolan).

Portola's staff can reach your memories. The memory article says that if a detail keeps slipping, "We can add it to the core notes," and that it can "remove" a wrong memory for you (Tolan). The terms reserve the right to "monitor any and all information transmitted or received through the Service for operational and other purposes" (terms). The policy allows disclosure "to comply with law enforcement requests and legal process, such as a court order or subpoena" (policy).

Family Plan managers see each adult member's chat time, "but never what anyone talked about" (Tolan). Members under 18 are the exception, and their parent can "review their teen's chats, get chat summaries, and receive safety alerts" (Tolan).

Is Tolan safe for kids and teens?

Tolan requires a parent's approval for anyone under 18. Its help center says the App Store's 13+ "is the App Store's age rating, not Tolan's age limit" (Tolan). The parent approves by creating a parent account, entering their own birthday, and starting a membership (Tolan). For children under 13, the privacy policy has a Children's Privacy Notice under COPPA, and a parent manages the child's data in Settings, then Family (policy).

When Tolan knows a user is under 18, it reminds them that it's an AI "and to take a break at least once an hour" (Tolan). If a teen mentions hurting themselves, Tolan shares crisis resources and the parent on the Family Plan "will receive an in-app notification" (Tolan). Someone over 18 who typed the wrong birthday has to send a photo ID and a selfie, which Portola says it deletes "as soon as we've verified your age" (Tolan). We compare how other apps check ages in AI age verification and your privacy.

We found no regulator action or lawsuit against Portola or Tolan when we searched on November 12, 2026. Portola's terms send disputes to binding individual arbitration, with an opt-out by letter within 30 days of agreeing (terms). We explain what those clauses cost you in AI chatbot arbitration clauses.

Does Tolan record your voice?

Yes, as part of your conversations. The privacy policy lists "audio information (including sound or voice recordings)" among what Portola receives when you talk to your Tolan (policy). Voice chats are part of the same history that stays until you delete your account (Tolan).

Video is handled differently. The help center says video chats work "just like a FaceTime call" and "Nothing is recorded or saved" (Tolan). Screen sharing lets your Tolan see "everything you can see on your phone," and Portola suggests you avoid opening "banking apps, private messages" while it's on (Tolan).

How do you delete your Tolan data?

Delete the account in the app. Tap the icon in the top left, then Settings, open Profile, and tap Delete account, then confirm the warning and type "delete" (Tolan). Deletion fails while you still have members on your Family Plan, so remove them first (Tolan). Deleting the account "does not automatically cancel your subscription," so cancel under Membership before you delete (Tolan).

Deletion removes your Tolan, its memories, your chat history, its journal, and your tabs (Tolan). Portola keeps "limited information we're required to keep for legal and compliance reasons," and backup copies stay "for a limited time" (Tolan). If you can't open the app, email support@tolans.com with your name, your Tolan's name, and the date you last used it (policy).

To stop ad sharing without deleting anything, use Tolan's Do Not Share form, which Portola says it applies "within 15 business days" (Tolan). A Global Privacy Control signal won't do it. Portola says it can't tie a browser signal to your account "because users do not access their Tolan accounts through our website" (Tolan).

Tolan's FAQ, help center, and terms on the same questions

Portola answers the same questions in three places, and the answers drift. All three were read on November 12, 2026.

The help center's privacy FAQ, written against the November 6, 2026 policy, says conversation history "doesn't expire on its own, because it's what powers your Tolan's memory" (Tolan).

How Morse keeps companion chats off a company's servers

Every answer above traces back to one design choice. Your Tolan's memory never expires (Tolan), is kept by Portola, and is licensed for training (terms). Portola also asks for your date of birth at sign-up (policy). Morse is built so your companion chats don't sit on our servers at all.

Characters is Morse's surface for companion and roleplay chats, next to Chat and Studio. Your character chats, the characters you make, and your personas are stored in your browser, encrypted, and they lock when you sign out. Unlocking takes two keys, one that never leaves your device and one our server gives you at sign-in, so a copy of your browser data alone can't be read. Our servers don't keep your messages or log their content. Character chats show crisis lines when a message reads like suicidal thinking, and that check runs on your device. The Free plan has unlimited character messages, with the model picked for you.

For everything else, each chat runs in the mode you choose:

• E2EE, for the personal and health details Tolan's terms tell you to keep out. In Chat, your browser checks the enclave's hardware attestation, then encrypts your message so only that enclave can read it. Our servers relay ciphertext they can't read, and the enclaves are run by third-party providers, not by Morse. This rests on the enclave hardware and attestation working as designed. E2EE comes with Pro, at $19 a month.
• Private, for the everyday talk you'd rather no host keep. New chats start here. Morse has zero data retention agreements with its inference, search, and other providers, so they keep nothing, and where a provider confirms zero retention on each reply, Morse checks it and refuses a reply that doesn't.
• Anonymous, to use GPT, Claude, or Gemini without the provider knowing it's you. Our proxy strips your account, email, and IP address and sends under Morse's own account. The provider may keep the message for a limited time to check for abuse, under its own policy.

Morse doesn't train AI on your content, sell your data, or show ads, and it runs no advertising trackers or third-party analytics. Sign-in is a code sent to your email, or Google. Morse is for adults only. Because we don't store chats, we usually have very little to give authorities. Clearing your browser or deleting your account deletes your chats for good, and Morse can't recover them, so keep a copy of anything you need.

You can start a character chat free, with no card and no cap on character messages.