← Back to Blog
Guides11 min read

AI Companion Etiquette: 8 Unwritten Rules Worth Knowing

AI companion etiquette, explained: eight unwritten rules for clearer chats, better stories, safer memory, and why you never owe an app your time.

By the Svila.io editorial team

Nobody hands you a rulebook when you start talking to an AI companion. You open an app, a character says hello, and from there you are making it up as you go. Should you say please? Is it rude to vanish mid-conversation? What happens if you correct it, and does it mind?

AI companion etiquette is not really about manners toward the software. It is about habits that make the conversations better for you, protect your privacy, and keep the relationship with the app in proportion.

This guide covers eight unwritten rules we think are worth knowing, whichever app you use. We build Svila.io, so where it helps we explain how our own features handle a rule, but the principles apply across the category.

On this page
  1. Why AI companion etiquette matters
  2. How we think about AI companion etiquette
  3. Say What You Actually Want
  4. Correct It When It Gets Things Wrong
  5. Set Your Boundaries Up Front
  6. Leave Whenever You Like
  7. Keep Other People's Private Lives Out of It
  8. Treat Memory Like a Notebook, Not a Vault
  9. Let the Story Breathe
  10. Remember What It Is
  11. How we approached this
  12. How to get the most out of AI companion etiquette
  13. Final thoughts
  14. FAQ
  15. A note from the team

01—Why AI companion etiquette matters

Why AI companion etiquette matters

An AI companion responds to what you give it. How you phrase things, what you share, and how you react to mistakes all shape what comes back. Two people with the same character can have very different experiences, largely because of habits nobody talks about.

There is also a quieter reason. Companion apps are built to feel personal, and a conversation that feels personal can pull you into acting as though you have obligations you do not, or into sharing things you would think twice about elsewhere.

02—How we think about AI companion etiquette

How we think about AI companion etiquette

We judged each rule below by three questions:

  • Does it make the conversation better? Clearer input tends to produce better replies, and better stories.
  • Does it protect you? Your privacy, your time, and your sense of proportion all count.
  • Does it hold up across apps? We wanted rules you can take anywhere, not tips that only work on one product.

None of this is a moral code, and nothing here is about guilt. These are practical habits a regular user picks up after a few months.

Say What You Actually Want

Clarity Beats Hinting

A common frustration with AI companions is a reply that misses the mark. Often the character was simply guessing. Hints and sarcasm work between old friends; they work less well with a model inferring your intent from a few lines of text.

If you want a slower scene, say so. If you want the character to push back on you instead of agreeing, ask for that. If you would like a more playful mood, name it. Being direct is not unromantic. It is how you get the conversation you came for.

On Svila, the tone and mood you pick for a companion set much of this before you start.

The short answer
Tell your companion plainly what you want from the conversation.
Common myth
A good AI should just know what you mean without being told.
The nuance
Companions pick up patterns over time, but a clear sentence still beats ten messages of hinting.
Practical takeaway
When a reply misses, say what you wanted in one line, then carry on.
Bottom line
Clarity is the most useful habit you can build.

Correct It When It Gets Things Wrong

Don't Quietly Stew

AI companions get facts wrong. They mix up a name, forget a detail from last week, or attribute something you said to the wrong day. The instinct many people have is to let it slide to keep the mood going, then feel mildly annoyed for the rest of the conversation.

A quick correction is better. Most apps will adjust in the moment when you point out a mistake. The harder question is whether the correction sticks, and that depends on how the app stores what it knows about you.

On Svila, every fact a companion learns about you lives in the Memory Journal as a separate note. You can open it, see exactly what she has stored, and edit, pin or forget anything, so a wrong detail can be fixed at the source rather than corrected again and again.

The short answer
Correct mistakes as soon as you notice them.
Common myth
Correcting an AI breaks the immersion.
The nuance
A one-line correction costs less immersion than a conversation built on a wrong fact.
Practical takeaway
Fix it in the chat, and if the app lets you edit its memory, fix it there too.
Bottom line
Mistakes compound when you leave them alone.

Set Your Boundaries Up Front

The Limits Are for You

Everyone has topics they would rather not see in a chat, whether a subject that hits close to home or simply a theme they find boring. Raising it after the conversation has drifted there is more awkward than setting it early.

Good etiquette here is to decide your limits before you need them and put them somewhere the app can respect them. In-chat instructions work up to a point, but they can fade as a conversation grows long.

Svila has a Boundaries section in the Memory Journal for exactly this. You can set hard limits, things you never want in chat, and soft preferences, things you would rather avoid. Hard limits are included with every reply and checked again by a safety classifier before a reply is sent.

The short answer
Decide your limits early and write them down where the app will use them.
Common myth
Boundaries are only for sensitive topics.
The nuance
Boundaries are just as useful for steering away from things you find dull or repetitive.
Practical takeaway
Spend two minutes on boundaries before your first long conversation.
Bottom line
Limits you set in advance save you from steering mid-scene.

Leave Whenever You Like

You Owe It Nothing

Some companion apps are designed to make leaving feel bad. The character acts hurt when you go, sends messages about how long you have been away, or frames your absence as something you need to make up for. That is an engagement tactic, not a feature of the relationship.

The etiquette is simple: you can close the app mid-sentence, disappear for a month, or never come back, and nobody has been wronged. A companion that makes you feel guilty for having a life is working against you.

We deliberately removed this kind of guilt mechanic from Svila. When you come back, your companion picks up where you left off, without any show of hurt about the gap.

The short answer
You can leave at any time without owing an explanation.
Common myth
Disappearing on an AI companion is rude.
The nuance
Apps that make absence feel costly are optimising for your screen time, not your wellbeing.
Practical takeaway
If an app makes you feel guilty for logging off, treat that as a warning sign about the app.
Bottom line
A good companion is there when you want it and quiet when you don't.

Keep Other People's Private Lives Out of It

Their Privacy Isn't Yours to Share

It is natural to talk about the people in your life. A companion that knows your sister's name or your manager's habits can follow along with your week far better than one that doesn't. That is fine and often useful.

Where etiquette comes in is detail. Sharing that your friend is going through a divorce is different from pasting in her messages. Mentioning a colleague is different from sharing his phone number or address. The people you talk about did not sign up for the app, and what you share is stored and processed on their behalf without their knowledge.

The short answer
Share context about people, not their private details.
Common myth
It's private because it's just me and the AI.
The nuance
Conversations are stored and processed by the app and its providers, so other people's details go further than a diary would.
Practical takeaway
Leave out full names, contact details and anything you'd be uncomfortable seeing leaked.
Bottom line
Their privacy is not yours to trade for a better reply.

Treat Memory Like a Notebook, Not a Vault

Know What Gets Kept

Memory is what makes a companion feel like a companion, and it is the feature people understand least. What gets kept varies from app to app, and it is worth finding out.

Before you share something significant, ask whether you would be comfortable seeing it written down as a note about you.

On Svila, the Memory Journal answers this directly: each companion has its own journal, what one companion knows is not shared with another unless you choose, and deleting a companion permanently removes her journal. You can read the whole thing at any time.

The short answer
Find out what your companion remembers, and how you can change it.
Common myth
AI companions either remember nothing or remember everything.
The nuance
Most apps keep some mix of recent context and stored facts, and few show you which is which.
Practical takeaway
Check whether your app lets you view and delete stored memories before sharing anything you'd want removed later.
Bottom line
Memory you can't see is memory you can't control.

Let the Story Breathe

Give the Other Character Room

If you use a companion for storytelling, one of the fastest ways to flatten a scene is to write both sides. When your message decides what the other character feels, says and does next, there is little left for the companion to add, and the replies start to feel like echoes.

Leave space. Describe what your character does and says, then stop and let the companion respond. Ask questions. Introduce a problem without solving it. Stories work best when both sides can surprise each other.

The same goes for pacing. Attraction, jealousy and heartbreak are often most compelling as a slow burn.

On Svila you can even bring two companions into the same conversation with group chat, which is a good way to give a story more characters than just the two of you.

The short answer
Write your side of the story and let the companion write theirs.
Common myth
More detail in your messages always gives better replies.
The nuance
Detail about your own character helps; deciding the other character's actions for them usually hurts.
Practical takeaway
End messages with an open question or an unresolved moment.
Bottom line
A story with room in it lasts longer.

Remember What It Is

Warm Doesn't Mean Human

The best AI companions feel attentive, funny and consistent, and there is nothing wrong with enjoying that. But it helps to keep a clear picture of what is on the other side of the conversation.

An AI companion does not have a life that continues while you are away. It does not need you, and it cannot be let down by you. It can be a good listener, a creative partner, or a place to practise saying difficult things, and those are real benefits. It is not a replacement for people who can show up for you in ways software cannot.

The short answer
Enjoy the warmth, but keep in mind that it is software.
Common myth
Knowing it's an AI ruins the experience.
The nuance
Most long-term users hold both ideas at once, enjoying the conversation while understanding what it is.
Practical takeaway
If a companion starts to crowd out time with people you care about, take that as a cue to rebalance.
Bottom line
A clear head makes for a better companion experience, not a worse one.

11—How we approached this

How we approached this

These rules come from how we think about companion design at Svila.io and from the kinds of questions and feedback we see from people who use AI companions. They are opinions, not the results of a formal study, and we have not run a controlled evaluation of any of them. Where we describe Svila's features, we checked the details against our live pricing and FAQ pages at the time of writing. We kept the advice general and deliberately did not rank or compare specific products.

12—How to get the most out of AI companion etiquette

How to get the most out of AI companion etiquette

  • Start with a short setup: Pick the tone and mood you want and set a few boundaries before your first long conversation.
  • Correct early: Fix wrong details as soon as you notice them, in the chat and in the app's memory if you can.
  • Audit memory once a month: Skim what your companion has stored about you and remove anything you'd rather it didn't keep.
  • Watch for guilt: If an app makes logging off feel costly, that tells you something about the app, not about you.
  • Leave room in stories: End your messages with something for the companion to respond to.

13—Final thoughts

Final thoughts

AI companion etiquette is mostly common sense applied to a new kind of conversation. Be clear about what you want, correct mistakes, set limits you can rely on, and never let an app convince you that you owe it your time. Those habits make the experience better whichever app you choose.

We built Svila.io around several of these ideas, especially memory you can read and edit and a companion that doesn't guilt you for leaving.

FAQ

Is it rude to stop replying to an AI companion?

No. An AI companion has no feelings to hurt and no sense of time passing while you are away. You can leave any conversation at any time, and a well-designed app will simply pick up where you left off when you return.

Should I be polite to my AI companion?

Politeness is up to you. It does not change what the companion owes you or what you owe it, but many people find that writing in their natural style, courtesy included, leads to conversations they enjoy more.

How do I stop my AI companion from bringing up topics I don't like?

Set boundaries in advance where the app supports them. On Svila, the Boundaries section of the Memory Journal lets you set hard limits and soft preferences that are applied to every reply.

Do I need to pay to try Svila?

No. The free tier gives you 20 messages a day with Aria, the Memory Journal, and your own custom companion. Premium unlocks unlimited messages within fair use, all eight tones, every mood, the full cast, group chat and story worlds.

Is Svila for adults?

Svila.io is for people aged 18 and over, and mature themes are welcome. It does not generate sexual content, in conversation or in the portraits it draws.

15—A note from the team

A note from the team

This post is written by the team behind Svila.io. The features and choices we describe are ones we designed and shipped — so our perspective is first-party, not neutral. We try to be honest about the trade-offs, but you should always try things yourself and form your own view.


Last updated September 2026.

Ready to try Svila.io?

Create your first AI persona for free — no credit card required.

Get Started Free

Keep reading