The short version
Habits gives you three main ways to control optional AI:
- Turn AI off completely with the master AI control.
- Manage individual data categories if you want AI in some parts of Habits but not others.
- Review the AI access history to see recorded AI data-access events by time, category and feature.
Core tracking remains available if optional AI is declined or disabled.
The nine category controls govern documented Habits tracking categories. Connected-service data such as Apple Health metrics or Google Calendar context needs separate verification because those are not named among the nine categories.
Habits currently uses OpenAI models for optional AI features.
AI requires a choice before optional AI processing
Habits includes an AI consent step.
You can:
- accept optional AI processing; or
- decline it.
Declining optional AI does not remove the core Habits trackers.
The exact trigger timing is currently a product-source conflict
The supplied product reference says both:
- AI consent is shown on first AI use; and
- a first-login consent modal is shown to every new user.
Those statements conflict.
This guide therefore does not publish an exact trigger point until the running build is checked.
The non-negotiable rule is:
Required permission must be obtained before personal data is shared for optional third-party AI processing.
Third-party AI disclosure
Habits currently uses OpenAI models for optional AI features.
The production consent experience should clearly explain that relevant information may be processed by the third-party AI provider before that processing occurs.
Open AI Privacy Dashboard from Settings
The current product places the AI controls in:
Settings → AI Privacy Dashboard
From there, you can manage the main AI switch and individual AI data categories.
Use the master AI switch if you do not want optional AI processing
The AI Privacy Dashboard includes a master AI on/off control.
Turning the master control off is intended to disable optional AI data processing.
Core tracking can still be used without optional AI.
That includes core areas such as:
- Medication tracking;
- Hydration;
- manual Food logging;
- Exercise tracking;
- Mood;
- Sleep;
- Weight;
- Progress Photos.
Turning AI off is not the same as deleting historical data
This guide does not claim that switching the master AI control off automatically deletes:
- previous AI chat history;
- previous AI access-log entries;
- existing reports;
- health records;
- processor records.
Use the relevant deletion / clear / account controls for those separate actions where available.
AI controls are not all-or-nothing
The current AI Privacy Dashboard includes separate controls for nine categories:
- Medication
- Hydration
- Food
- Exercise
- Mood
- Weight
- Sleep
- Photos
- Measurements
These controls are intended to let you limit which categories are available to optional AI processing.
That means you do not have to treat:
AI on
as:
AI can use every category.
Important limitation
The exact effect of turning a category off can depend on the AI feature.
This guide does not claim that every AI feature maps one-to-one to one category until the current feature-permission mapping is verified.
AI controls are about optional AI use — not whether Habits can keep the underlying record
For example:
- turning the Medication AI category off does not delete your Medication records;
- turning Mood AI access off does not remove the Mood tracker;
- turning Photos AI access off does not mean your existing Progress Photos are automatically deleted.
The category controls are intended to limit optional AI use of those categories.
Core non-AI product functionality can still use the records you have chosen to keep in Habits.
This is also not a deletion control
If you want to:
- delete a health record;
- clear AI access history;
- delete an account;
use the relevant separate product control.
The nine category controls do not name Apple Health or Google Calendar
Where connected features are available, Habits can also work with information from services you choose to connect.
The current product reference documents:
- selected Apple Health metrics such as Steps, Heart Rate, HRV and Blood Oxygen;
- read-only Google Calendar context for the Pro Calendar integration.
The same product reference says:
- wearable metrics can be included in AI correlations where available;
- Calendar commitments can be used as context for the rollout-controlled Daily Plan.
But Wearables / Apple Health and Calendar are not named among the nine AI category controls.
So this guide does not claim that the nine category switches fully govern those connected-service data flows.
Use the connected-service controls too
Where applicable, you can also control connected services through:
- Apple Health permissions;
- the Google Calendar disconnect/revoke flow.
The final Privacy Policy should explain exactly how those connected-data permissions interact with optional AI.
“AI can access a category” and “AI can write to that tracker from chat” are different things
The current verified conversational AI write-back categories are:
- Hydration;
- Food;
- Mood;
- Weight;
- Exercise.
For example, supported chat input can create records in those trackers.
The current product reference does not verify conversational AI write-back for:
- Medication;
- Sleep;
- Photos;
- Measurements.
So a Medication AI permission does not mean this guide claims:
“Tell AI you took your dose and it will log it.”
The same distinction applies to other categories.
See recorded AI access events
The AI Privacy Dashboard includes an AI access history.
The current product reference says each recorded event can show:
- time;
- data category;
- which AI feature accessed it.
Current visible history
The current plan matrix documents:
- Free: last 5 entries visible;
- Pro: last 20 entries visible.
The underlying privacy controls themselves are available in both Free and Pro.
What the history is for
The log is useful for a practical question such as:
“Which part of Habits used an AI data category recently?”
The documented fields are event metadata:
- time;
- category;
- feature.
This guide does not claim that the access history shows:
- the exact health values used;
- the complete prompt sent to a model;
- the provider/model name per event;
- processor retention;
- every internal security log.
It should not be interpreted as a complete legal audit of every processor log or retention record outside the app.
You can clear the visible AI access history
The current product reference says the AI access history can be cleared.
Clearing that history is a separate action from:
- deleting your Habits account;
- deleting health-tracking records;
- deleting previous AI chat history;
- changing the master AI switch;
- changing category permissions.
Do not assume one privacy control performs all of those actions.
Habits documentation describes PII masking for supported AI processing
The current product reference describes PII masking that filters direct identifiers such as:
- name;
- phone number;
- health IDs.
The privacy-specific section explicitly names these supported flows:
- chat;
- insights;
- reports;
- Daily Plan generation.
PII masking does not mean anonymous
AI may still need relevant context from the tracker categories you have allowed in order to provide the AI feature.
For example, removing a name does not turn health information into anonymous data automatically.
So Habits should describe this as:
PII masking
not:
complete anonymisation.
Image limitation
This guide does not claim that text or identifying details visible inside an uploaded image are automatically removed unless that image-processing behaviour is separately verified.
Previous AI chats can remain in Habits
The current product stores persistent AI conversation history server-side.
That is separate from:
- the AI access history;
- the master AI control;
- category permissions.
Turning AI off controls optional AI processing going forward.
This guide does not claim that turning AI off automatically deletes previous AI chats.
The final Habits Privacy Policy should explain the approved retention and deletion treatment.
Habits currently uses OpenAI models for optional AI features
OpenAI is the current AI provider documented for Habits optional AI functionality.
That provider disclosure is separate from the question:
What does OpenAI retain or use?
The final Habits Privacy Policy should describe the actual Habits configuration and approved provider terms.
What this Product Guide does not promise
This guide does not say:
- OpenAI retains zero Habits data;
- Habits has Zero Data Retention enabled;
- OpenAI is technically incapable of retaining anything;
- Habits data is used for model training;
- Habits data is never used for model training.
Those answers depend on the actual Habits account/configuration and agreements.
The core tracker remains available
If you do not want optional AI, you can still use the core tracking system.
The current Free/Pro matrix documents core features such as:
- Medication tracking and reminders;
- Hydration tracking and reminders;
- manual Food logging;
- Exercise tracking and structured programmes;
- Mood;
- Sleep;
- Weight;
- Progress Photos;
- export;
- biometric lock;
outside the requirement to enable optional AI.
What you give up
Features that depend on optional AI will not work normally while AI processing is declined or disabled.
That can include supported functionality such as:
- AI chat;
- AI Food Photo;
- AI Workout generation;
- AI Insights;
- AI Nutritional Analysis;
- AI-generated planning/report functionality where applicable.
Privacy controls are not a Pro upgrade
The current Habits plan matrix includes the AI Privacy Dashboard in:
- Free;
- Pro.
The master AI control and category controls are not documented as Pro-only features.
The visible AI access-history window currently differs:
- Free: last 5 entries;
- Pro: last 20 entries.
The current master/category controls should not be positioned as a reason users must pay to protect their data.
The product currently shows a larger visible AI access-history window in Pro, which this guide discloses separately.
You can make the AI setup fit your preference
Here are examples of how the current control model is intended to work.
I do not want optional AI
Use the master AI control to turn optional AI processing off.
I want AI, but I want to limit some data categories
Keep the master AI control on and manage the individual category controls.
I want to see which category an AI feature accessed
Review the AI access history.
I want to remove the visible AI access history
Use the clear-history action.
Important
This guide does not claim that changing a category permission:
- deletes historical data;
- deletes previous AI chat;
- deletes processor records;
- removes that category from every historical report.
Those are separate data-retention/deletion questions.
AI privacy-control questions
Do I have to use AI in Habits?
No.
The current product allows optional AI to be declined while keeping the core trackers available.
Where are the AI controls?
The current product places them in:
Settings → AI Privacy Dashboard
Can I turn all AI off?
Yes.
The AI Privacy Dashboard includes a master AI control.
Can I choose which types of data AI can use?
The current Dashboard includes separate controls for:
- Medication;
- Hydration;
- Food;
- Exercise;
- Mood;
- Weight;
- Sleep;
- Photos;
- Measurements.
Does having a Medication AI permission mean chat can automatically log my Medication?
No such conversational Medication write-back is verified.
The current conversational write-back categories are Hydration, Food, Mood, Weight and Exercise.
Can I see when AI accessed a category?
The current AI access history records events with time, category and feature information.
How much access history can I see?
The current plan matrix documents:
- Free: last 5 entries;
- Pro: last 20 entries.
Can I clear the AI access history?
Yes.
The product reference says the visible AI access history can be cleared.
That does not mean every health record, prior AI chat or processor record is deleted.
Does turning AI off delete my previous chats?
This guide does not claim that.
The current product stores persistent AI chat history server-side, and the detailed retention/deletion behaviour belongs in the final Privacy Policy.
Does Habits remove my name before AI processing?
The current PII masking system is documented to filter direct identifiers including names, phone numbers and health IDs for supported AI processing.
That is not the same as making all health context anonymous.
Does Habits use OpenAI?
Yes.
OpenAI is the current AI provider documented for optional Habits AI features.
Is Habits AI anonymous?
No blanket anonymity claim is made.
PII masking can remove certain direct identifiers, while relevant health/tracking context may still be processed.
Are AI privacy controls Pro-only?
No.
The current plan matrix includes the AI Privacy Dashboard in Free and Pro.
Use AI where it helps. Keep control over where it fits.
Turn optional AI off completely, manage individual data categories, or review recorded AI access from the AI Privacy Dashboard.
Small print:
Habits currently uses OpenAI models for optional AI features. The Habits Privacy Policy explains the approved current data-handling arrangements.