Consumer Health Data Privacy Policy

For residents of Washington State, Nevada and Connecticut. Last updated: 8 November 2026

ORIZN Sober is run by ORIZN LTD (formerly SoberStreak AI Ltd).

This policy explains how ORIZN LTD ("ORIZN", "we") handles consumer health data, as defined by Washington's My Health My Data Act. It is separate from our general Privacy Policy, which also applies.

Nevada and Connecticut have consumer health data laws similar to Washington's. If you live in either state, you have the same rights set out below and can exercise them in the same way. Where those laws give you anything Washington's does not, you have that too.

What we collect

ORIZN Sober is a recovery tracking app. The information you give us is consumer health data under the Act, and includes:

What you are recovering from, and when you started
Your streak, and any slips you record
Urges you log, and how you rated them
Journal entries you write
Your conversations with the AI Companion
If you use these Premium features (in ORIZN Sober 1.1.5 and later): the journal reflections you choose to keep, your weekly notes, and the short summary kept by Companion memory
If you write plans for your triggers (Premium; in ORIZN Sober 1.1.5 and later): what you write in them, and the name and number of anyone you add as someone to text
If you plan for a tough day (Premium; in ORIZN Sober 1.1.5 and later): the kind of day or your title for it, its date and time, your plan, the name and number of anyone you add as someone to text, and how the day went if you tell us
If you set savings goals (in ORIZN Sober 1.1.5 and later): each goal's name, target and, if you set one, the date you'd like to reach it by, and which recovery it follows
If you choose what your savings stand for (Premium; in ORIZN Sober 1.1.8 and later): each thing's name and the price you type
If you use the morning intention (Premium; in ORIZN Sober 1.1.5 and later): the line you write for each day
If you follow a guided programme (Premium after the first 3 days; in ORIZN Sober 1.1.5 and later): when you started and which days you have completed
Milestones you reach
If you use the Community: what you post, who you follow, and your messages

We collect this only from you. We do not buy health data, obtain it from data brokers, or infer it from your activity elsewhere.

Why we collect it

To run the app you asked us to run: to track your recovery, show your progress, let you write and look back on your journal, answer you through the AI Companion, provide the Premium features you turn on (journal reflections, weekly notes and Companion memory, in ORIZN Sober 1.1.5 and later), and — if you use it — let you take part in the Community. We also use it to keep the service safe, which includes detecting content that suggests someone is in crisis.

We do not use it for advertising, and we do not profile you for marketing.

Who we share it with

We do not sell your health data. Not for money, and not for anything else of value. We have never done so and have no plans to.

We share it only with service providers who help us run the app, under contract, and only for that purpose:

Who	What they receive	Where
Google (Firebase)	All app data, stored and processed	Your recovery data, journal, Companion conversations and Community content are stored in London, United Kingdom. Photos and journal voice notes you upload are stored in Turin, Italy. Your sign-in details (email address and password) are held by Google's authentication service in the United States.
OpenAI	Your AI Companion messages, to check them for signs of crisis. Before 8 November 2026 OpenAI also wrote the Companion's replies; Anthropic writes them now (see below). Before 8 November 2026, if Personalise Companion was on: your name, streak and start dates, slips, plan answers, recent journal excerpts, urge counts and today's morning intention. Our privacy policy lists exactly what. Whether or not it is on, a chat you start from a milestone or a weekly note includes that milestone's day count or that note, and a saved chat you continue sends its earlier messages again. Before 8 November 2026, if you chatted with Rae during set-up: that one message, what you are recovering from, your start date and streak, your daily spend and currency, your hard moments and what you want back, whether or not Personalise Companion is on. With Premium (ORIZN Sober 1.1.5 and later), only for features you use: a journal entry and its mood when you tap "Reflect on this", to check it for signs of crisis and write a reflection; that week's journal entries and urge and slip notes if weekly notes are on, to check them for signs of crisis (the note itself is written from figures, never your words); the messages you sent in each conversation if Companion memory is on, to write a short summary that is then sent as part of later conversations to the provider writing the replies, except any conversation that looked like a crisis, which is never remembered; and the trigger or kind of day (or your title for it) and the steps you have written if you ask for help writing a plan, never the person you add to text and never the date. Community posts, messages and photos, to check them for safety. Never your email address or account identifier. Never used to train AI models.	United States
Anthropic	Since 8 November 2026: your AI Companion messages, to generate replies, plus the details about how you use the app, and your set-up chat and answers if you chat during set-up. If Personalise Companion is on, the information listed for it above. If Companion memory is on, your summary. Crisis checks and photo moderation remain with OpenAI. What we send Anthropic is covered by its published terms for API customers like us. It deletes what we send and what it writes back from its backend within 30 days of receiving or generating it, with exceptions that its retention page lists. It does not use them to train its models unless the customer (that is us) explicitly reports material to it or opts in, and we do neither. We don't control what its systems flag. If Anthropic's automated trust and safety systems flag content as violating its Usage Policy, it keeps those messages and replies for up to 2 years, and the classification scores for up to 7 years.	United States
Resend	Crisis alert emails — a post reference and a timestamp, never your words	United States
RevenueCat	Subscription status only — no health data	United States

If you use the Community, what you post there is visible to other members. That is you sharing it, not us.

One thing worth calling out, because you do not post it and it still appears: if you have set a profile photo, a copy of it is shown next to what you post in the Community. You never have to set one, and members who have not set one are shown their initials instead. Removing your photo removes the community copy with it. Your Community display name is separate and is only ever what you type yourself—we never take it from your account name.

Your rights

If you live in Washington, Nevada or Connecticut, you may:

See what we hold about you. Email us and we will tell you, including a list of the third parties above.
Delete it. You can delete your account in Settings (Profile › Account › Delete Account), which deletes your data from the systems we run on Google Firebase, subject to the short backup and recovery periods described below. You can also email us. When you delete your account we also pass the deletion on to our service providers. RevenueCat and Resend hold no health data about you. We delete your subscription record with RevenueCat when your account is deleted. Resend keeps a log of the emails we have sent you for a limited time under its own retention policy, and holds nothing else about you. We never send OpenAI your email address or account identifier, but what it holds can still identify you: what you write may say who you are, and before 8 November 2026 it received your name if Personalise Companion was on. It deletes everything we send within 30 days as a matter of course, unless the law requires it to keep something longer. What OpenAI receives is set out in the table above. Anthropic's published terms are to delete what we send it within 30 days, with exceptions that its retention page lists. It keeps content its automated safety systems flag for longer, for the periods in the table above, and we cannot delete that content earlier, including when you delete your account. Backup copies of our database are kept for up to 14 days after deletion and then expire, and a backup is never restored in a way that brings back a deleted account. Deleted photos and voice notes can remain in our storage provider's recovery area for up to 7 days before they are gone for good. One exception is yours to choose: if, when deleting your account, you ask us to leave your Community posts and replies up, we take your name, photo and profile link off them, and the words you wrote stay as you wrote them.
Withdraw your consent to us collecting and sharing your health data. Because the app cannot work without processing this information, withdrawing means closing your account.

We will not treat you differently for exercising any of these rights.

To make a request, email hello@soberstreak.ai from the email address on your account. We will respond within 45 days, and will tell you if we need longer.

If we refuse a request, we will tell you why, and you can appeal by replying to that email. If we refuse your appeal you may contact the Attorney General of your state.

Location

We do not use geofencing, and we do not collect or infer your precise location.

Contact

ORIZN LTD
hello@soberstreak.ai