How AI Chat & Smart Logging work in Habits

The short version

Habits AI Chat can do two different jobs:

Conversation

You can ask questions or interact with the AI interface.

Smart Logging

For currently supported tracker types, Habits can extract structured information from what you say and create a tracker entry.

The verified write-back set is:

• Hydration;
• Food;
• Mood;
• Weight;
• Exercise.

Smart Logging does not currently verify conversational write-back for Medication, Sleep, Photos or Measurements.

AI Chat is linked from the Habits Today experience

The current Today dashboard includes an AI Assistant card that links to the AI chat screen.

The current app also stores previous AI conversations so the chat can persist across sessions.

Habits can suggest questions based on available context

The product source documents AI-generated suggested questions when Chat loads.

Those suggestions can be contextual to health/tracking information available to the AI feature.

A suggested question is:

a prompt to start a conversation

not:

a medical recommendation.

You do not need to use a suggested question.

Supported chat messages can become tracker entries

The current product source documents Smart Logging through conversational extraction.

For example, information typed into Chat can be turned into a supported tracker record rather than making you open each tracker manually.

Five verified write-back areas

Habits currently verifies chat write-back for:

• Hydration;
• Food;
• Mood;
• Weight;
• Exercise.

A tracker entry is different from ordinary conversation

Not every sentence should become a health record.

Habits should only create a record when the conversation contains enough supported information to do so.

Chat can create a Hydration entry

The source gives a Hydration example as a verified Smart Logging use case.

If you describe a supported Water/fluid entry clearly enough, Chat can create the corresponding Hydration record.

Exact quantities matter

A phrase such as:

“two glasses of water”

does not inherently specify an exact millilitre/ounce volume.

This guide does not claim Habits uses a particular glass-size conversion until the running extraction rules are verified.

Chat can create a Food entry

Food is one of the verified conversational write-back categories.

The broader Habits Food system stores:

• meal name;
• meal type;
• calories;
• Protein;
• carbohydrates;
• fat.

AI-assisted Food logging can involve estimates.

That means:

an AI-created Food log should not be treated as measured nutrition.

Chat can create a Mood entry

Mood is one of the verified write-back categories.

The Habits Mood tracker uses a five-point scale:

• Awful;
• Bad;
• Okay;
• Good;
• Great.

Conversation is not automatically a Mood log

A user may discuss feelings without intending to record a Mood entry.

The product should distinguish:

talking about Mood

from:

asking Habits to log Mood.

Chat can create a Weight entry

Weight is one of the verified Smart Logging categories.

Habits supports:

• kg;
• lb.

When logging Weight through conversation, the unit and value need to be clear.

Chat can create an Exercise entry

Exercise is one of the verified conversational write-back categories.

A supported message can create an Exercise record in Habits.

Core tracking remains available if you do not want AI Chat

Habits includes an AI consent system.

If optional AI is declined or disabled:

• AI Chat should not process a new conversation;
• core trackers remain available.

Habits currently uses OpenAI models for optional AI features.

Chat should follow the AI data permissions you choose

Habits' AI Privacy Dashboard includes category controls for:

• Medication;
• Hydration;
• Food;
• Exercise;
• Mood;
• Weight;
• Sleep;
• Photos;
• Measurements.

Those controls govern optional AI access.

The five verified conversational write-back categories are a smaller set.

Learn more: How AI Privacy Controls Work in Habits →

Habits records AI data-access events

The AI Privacy Dashboard includes an AI access history with:

• time;
• data category;
• AI feature.

Chat is one of the supported AI flows.

Smart Logging audit question

If one message writes to:

• Food;
• Hydration;

the access history should accurately represent the categories used.

Access audit ≠ mutation audit

The documented AI access history records:

• time;
• category;
• feature.

That does not necessarily tell the user:

“AI created a 500 ml Hydration record.”

For Smart Logging, consider a separate product-visible action acknowledgement/history containing:

• what tracker was changed;
• what value was written;
• time;
• Undo/Edit where applicable.

Do not overstate the AI access log as proof of every record mutation.

Habits documentation specifically includes Chat in its direct-identifier masking flow

The current Privacy source says PII masking applies to Chat.

It is documented to remove direct identifiers such as:

• name;
• phone number;
• health IDs;

before supported content reaches the AI model.

Masking is not anonymity

The Chat may still need relevant permitted health/tracking context to respond.

Removing a name does not make the remaining health information anonymous.

Your AI conversation can persist across sessions

The current product source says AI conversation history is stored server-side.

That allows previous Chat conversation history to persist.

Turning AI off should not be interpreted as automatically deleting earlier Chat history.

The final Privacy Policy should explain the verified retention and deletion treatment.

AI Chat is included in Free and Pro

The current plan matrix includes AI Health Chat in:

• Free;
• Pro.

Both plans have a daily AI allowance.

Pro has a higher allowance.

Exact usage numbers

This guide does not publish exact call counts until customer-facing quota semantics are finalised.