Safety on Chatili

Anonymity is your first line of defense

You stay safe on this platform mostly because nobody can identify you. The profile is four fields and none of them are tied to a real-world identity. The chats are not stored after 24 hours. There is no profile photo, no public username, no friend list, no shareable link to "your account". A bad actor cannot pivot from your conversation to anything else about you — unless you give them the way to.

That is the one rule that does more for your safety than anything we can build: do not share information that de-anonymises you. Real name, address, phone number, social-media handle, school, employer, photo of your face, video of your room. Country is fine (it is already on your profile). Everything else is yours to keep.

You choose who you talk to

The directory model is itself a safety feature. Instead of being thrown into a chat with whoever the system happened to pair you with, you see who is online and pick. If a card looks off, skip it. If a conversation starts off badly, leave — there is no penalty, no streak, no "you have skipped too many users" message.

You can also filter the directory by country, gender, and age range to narrow the pool to the kind of conversations you are actually here for.

Block: the instant escape hatch

Inside every conversation, the chat header has a Block button. Tapping it does three things at once: it ends the current conversation, it removes the chat from your inbox, and it tells the server to refuse any future message from that session to you. Block is one-way and immediate. You owe no warning, no "I am about to block you" message.

Use Block generously. The cost of blocking someone you would not have minded talking to is zero — there are thousands of other online users — and the cost of NOT blocking someone who is making the chat unsafe is real.

Report: the moderation feedback loop

Report is the right tool when the behaviour is specifically harmful, not just unwanted: harassment, hate speech, sexual content directed at you, anything illegal, anything involving someone who appears to be a minor. Report sends the session id and your reason to an internal moderation log.

Three independent reports against the same session within a rolling seven-day window automatically remove that session from the platform. The session is kicked from the directory, the WebSocket is closed, and a "blocked" flag is set against the session id so it cannot reconnect for the remainder of its lifetime. No human review is required — the rule is by design.

Automated guardrails

Two systems run silently in the background. The first is a per-session rate limiter: each session can send up to five messages a second, with separate caps on block, report, directory queries, and chat openings. The second is a behavioural bot detector that watches for non-human patterns (uniform inter-message timing, instant messaging on join, mass chat-opening, content that does not vary). Sessions that look like bots get shadow-banned: they continue to operate from their own perspective but their messages are never delivered.

These guardrails are why the platform stays usable even though nobody is reading your chats.

Why we do not read your chats

A human moderator reading anonymous chats does more harm than good. They cannot scale (millions of messages, tens of moderators), they create a privacy hole (an employee with database access to "anonymous" conversations), and they make false-positive removals that punish the wrong people. We chose structural safety — Block, Report, rate limits, bot detection, automatic kickout — over content monitoring.

The trade-off is that this only works if you USE Block and Report when you need them. The platform protects you when you tell it to.

Signals that a conversation is going wrong

The most reliable early-warning signs of a scam, a manipulation attempt, or a predatory user: the other party tries to move the conversation off the platform within the first few minutes ("send me your Snap", "let's switch to Telegram"); they ask for photos or video; they get unusually intimate unusually fast; they ask leading questions about your age or location; they push a link, a wallet address, an investment opportunity, or a job that "anyone can do from home".

Treat any of these as an automatic Block. None of them are normal in a stranger chat that exists for conversation.

Zero tolerance for conduct involving minors

The platform is for adults aged 18 and above. Any attempt to engage someone presenting as a minor in sexual conversation, solicit images, or share content depicting the abuse of a minor is grounds for immediate permanent removal and, where required by law, preservation of session metadata for transmission to law-enforcement and child-safety authorities. Use Report and the platform handles the removal.

Using the platform on a shared device

Because your session is kept in browser local storage to support the "close the tab and come back" flow, anyone with physical access to your browser profile can resume your session for the next 15 minutes. If you are on a shared device, click "End session" before walking away, or use an incognito window (which clears your storage automatically when you close it).

When something has already gone wrong

If you shared something you regret, change any password you may have handed over, watch your accounts for unusual activity, and block any further contact. If money was sent or fraud is suspected, contact your bank or payment provider immediately — many fraud transactions are reversible if reported within 24 hours.

If a stranger sent you content that is illegal in your jurisdiction (e.g. credible threats, child sexual abuse material), screenshot the chat window before closing it (we do not retain message content, so a screenshot is the only durable evidence), then use Report and contact local law enforcement.

Age requirement

Chatili is restricted to users aged 18 and above. Misrepresenting age is grounds for removal.