Privacy
Last updated: November 16, 2026
Operated by: Together with Kai, Inc.
Together with Kai, Inc. ("Together with Kai," "we," "us," or "our") respects your privacy and is committed to protecting the personal data of our users ("you," "user," or "member"). This Privacy Policy describes how we collect, use, process, disclose, and safeguard your personal information when you access or use our website, mobile applications (iOS and Android), communication tools, and Kai, our artificial intelligence companion.
Please read this Privacy Policy carefully. By creating an account, accessing our services, or interacting with Kai, you acknowledge that you have read, understood, and agreed to the practices described in this policy.
A. Information You Provide Directly
Account Information
Name, email address, chosen username, password credentials, and mobile phone number (if you sign up with it, or opt in to transactional or conversational communications via WhatsApp).
Profile Details
Optional demographic and personal information you choose to provide, including last name, gender, birth year, city, country, bio, interests, and profile pictures.
Conversations with Kai
User prompts, text messages, conversational history, voice recordings, uploaded visual media, and user-generated creative prompts exchanged with Kai.
Community Interactions
Messages sent in Circles, direct communications between members (including voice notes, which we transcribe so they can be shown as text), user reactions, comments, and posts in Spaces.
B. Audio and Video Interactions (Spark)
Live Communications
Spark enables one-on-one audio and video connections between consenting users.
Session Processing and Safety Sampling
We do not record or retain continuous full audio or video streams. To enforce community safety standards and prevent abuse, our automated safety systems analyze short real-time audio transcripts and periodically capture visual safety snapshots (approximately once every 20 seconds).
Snapshot and Transcript Retention
Snapshots and transcripts from parts of a call with no safety alert or user report are automatically deleted within about 20 minutes. Snapshots and transcripts flagged by our automated abuse-detection systems, or associated with a user report, are retained in encrypted storage for up to 14 days for manual review, after which they are permanently deleted. Where a transcript led to a strike, that transcript is kept for up to 90 days so the decision can be reviewed or appealed; snapshots are never kept beyond 14 days.
C. Inferred Data, Behavioral Patterns, and Psychological Profiles
AI Inferences and Assessments
Based on your interactions with Kai, our algorithmic systems derive behavioral traits, conversational context, values, goals, connection tendencies, and personality profiles (such as an estimated Big Five-style trait index).
Non-Clinical Disclaimer
All personality metrics, emotional baselines, and compatibility ratings are programmatic algorithmic estimates designed solely for entertainment, engagement, and platform matching. They do not constitute psychological, psychiatric, medical, or clinical evaluations.
D. Automated Signals and Distress Detection
Safety & Crisis Signals
Our systems monitor textual messages and conversational context for potential indicators of distress, emergency, or imminent self-harm.
Limitation of Automated Safety
Distress detection is an automated software tool and cannot guarantee the identification of all emergencies or indicators of harm. Our service is not a medical provider, crisis intervention hotline, or emergency response service.
E. Device, Technical, and Transactional Data
Usage & Diagnostics
Device identifier, advertising identifiers (IDFA/GAID), IP address, browser type, operating system version, crash reports, and interaction logs.
IP Addresses, Security & Prevention of Abuse
We use your IP address to run the service, to stop spam, bots, fraud, and ban evasion, and to deliver and measure advertising. To stop banned members from creating new accounts, our website and apps also keep a random device code (in your browser and app storage) and send it to us when you sign up or sign in. We store only a one-way cryptographic hash of it, never the code itself and never hardware identifiers such as IMEI; hashes are kept for 90 days, or while an account linked to the device is banned. For sign-up abuse checks we store a one-way cryptographic hash of your IP address for about 24 hours; hashes linked to banned accounts are kept for 30 days. Your IP address is also stored with some sign-up, phone verification, and message delivery records for security and troubleshooting, and is shared with Thrad and Meta as described in Section 7.
Payment Data
Subscription billing is handled entirely by certified third-party payment processors (e.g., Stripe, RevenueCat, Apple, and Google). We receive confirmation of active subscription status and renewal events, but never store or access full payment card numbers.
We process personal data for specific, lawful purposes:
Public Information
Your username, first name, profile picture, bio, and designated public profile answers are visible across the platform, including to people who are not signed in.
Standard Member Visibility
Signed-in members may view your last name, calculated age, gender, city, country, and interests, unless restricted.
Private Mode
Enabling Private mode in Settings hides your last name, age, gender, and city from other members. Your username, first name, bio, country, and selected interests remain visible to facilitate compatibility matching. Note: Private mode governs member-to-member visibility and does not alter programmatic data processing required for ad delivery or service operation.
Memory Management
You may reset Kai's conversational memory or turn off Kai chat at any time in Settings > Security & Privacy > AI Features. Resetting memory clears the remembered facts and profile Kai has built about you.
We do not sell your personal data for monetary consideration. We share information strictly under the following parameters:
Provider
Purpose & Data Scope
Supabase
Login, database, and backend hosting
Vercel
Website and app hosting and content delivery
OpenAI
Kai chat replies, summaries and memory; safety checks on public posts, profiles and chats; matching and conversation starters; voice transcription; picking ad topics (Section 7)
Google (Vertex AI / Gemini)
Photo and image checks; Spark and Circle safety checks; voice transcription, including voice notes in private messages; images you ask Kai to create
Google Web Risk
Checking links against known phishing and malware sites
Google Maps
Place search when you choose to add a location
GetStream
Member to member messaging, live audio and video, and moderation tools
Stripe and RevenueCat
Subscription payments
AiSensy
WhatsApp messages you opt in to
Resend and Klaviyo
Account and service emails
Firebase Cloud Messaging
Push notifications
WebToNative
The iPhone and Android app wrapper
Mixpanel, Google Analytics and Google Tag Manager
Usage analytics
Sentry
Error reporting and crash diagnostics
Cloudflare
Bot protection and web security
Thrad
Sponsored cards in Kai chat, after Spark calls, and in private messages, Circles, posts and image creation. See Section 7
Google AdSense, Google Ad Manager and Google AdMob
Advertising on our website and in our apps. See Section 7
Meta (Conversions API)
Measuring our own advertising campaigns when you sign up or subscribe. See Section 7
Each has its own privacy policy. OpenAI and Google process data under enterprise API terms and do not use it to train their models. Error reports sent to Sentry can include your account identifier.
Enterprise AI Safeguards
Text, audio, and visual data processed via commercial API partners (OpenAI and Google Vertex AI) are processed for programmatic inferencing and real-time moderation; our contractual agreements prohibit third-party model training on your private conversational inputs.
Exclusions from AI Processing
Account credentials, authentication passwords, cryptographic keys, and payment tokens are never exposed to AI model endpoints.
Private Message Review
Member-to-member direct messaging is protected in transit and at rest; it is not end-to-end encrypted. AI models do not monitor private member messages for moderation; protections are limited to an automated filter for a short list of prohibited words and checks that hide links known to be unsafe. Voice notes in private messages are transcribed by AI so they can be shown as text. If a message is reported by a recipient, our moderation staff will review the reported conversation.
To choose relevant ads, an AI model reads your own recent messages in a private chat (never the other person's) and picks out a few interest keywords and a one-line summary, such as "planning a trip to Goa". Only those keywords and that summary are used for ads; your messages themselves are never shared with advertisers. Nothing is created from a chat that shows distress or was recently flagged by our safety systems, from messages about health, feelings or other sensitive topics, or if you turn off Personalized Ads. These keywords are deleted after 14 days. See Section 7.
Your Consent
We ask for your permission the first time you open Kai chat. That permission covers Kai chat only. The other AI uses in this section are part of using Together with Kai and are covered by our Terms of Service, Section 12. Emotional and distress signals are used to support you and keep the community safe; they are never used to target ads.
A. Ad-Supported Platform
Free accounts may receive full-screen, banner, interstitial, video, and contextual advertisements. Paid subscribers do not receive advertising. Advertising never affects who you are matched with or what your feed shows you.
B. Advertising Data Sharing with Thrad
To provide contextual and demographically appropriate advertising cards in Kai chat, after Spark calls, and in private messages, Circles, posts and image creation, account data of members outside the European Economic Area, the United Kingdom, and Switzerland may be shared with our monetization partner, Thrad:
We decide whether you are in the European Economic Area, the United Kingdom, or Switzerland from the country your connection comes from or, when that is not available, the country on your profile. We never share your name, bio, gender, or anything Kai remembers about you.
Thrad uses this information only to choose and measure ads inside Together with Kai: to recognise you as the same person, to limit how often you see the same ad, to match you with audiences that advertisers provide in the same hashed form, and to build a profile of your interests from the keywords and summaries above. Thrad does not sell this information, does not keep the text of conversations, and deletes your identifiers and interest profile within 60 days of the last ad shown to you.
C. Google and other advertising partners
On our website, Google AdSense and Google Ad Manager show ads; in our apps, Google AdMob shows ads. Google receives your device's advertising identifier (in our apps), IP address, approximate location, and basic interaction and diagnostic data. If we add another advertising partner, we will update this policy before it receives any of your data.
D. Measuring our own advertising (Meta)
When you sign up or subscribe, we tell Meta that the event happened so we can measure our own advertising campaigns. We send a hashed form of your email address, phone number, first name, and country, together with your IP address, browser or device information, and Meta click identifiers.
E. Automated Distress Gating
Where our safety systems detect distress, emotional vulnerability, or crisis markers in a Kai conversation or a Spark call, that content is excluded from ad requests and no ad is shown after that Spark. You will never see an ad on a crisis screen. Disclaimer: while designed to prevent monetization during vulnerable moments, automated distress detection relies on software and cannot guarantee absolute filtering.
F. Ad Preference Controls
You may turn off "Personalized Ads" in Settings. When it is off, we stop sharing your hashed identifiers, age group, profile country and city, and interests with Thrad, we stop creating ad keywords from your private messages, and Google ads on our website switch to non-personalized ads. Ads are then matched only to the content on your screen. Your IP address and device information are still used to deliver the ad. Google AdMob ads inside our iPhone and Android apps do not yet follow this setting; you can limit them in your device's ad settings, and in our apps we ask your permission before your advertising identifier is used.
G. Cookies and Google advertising on our website
We use cookies that are essential (login, session, security), functional (preferences), analytics, and advertising. On our website we use Google AdSense. Third-party vendors, including Google, use cookies to serve ads based on your prior visits to our website or other websites. Google's use of advertising cookies enables it and its partners to serve ads to you based on your visit to our site and/or other sites on the internet. You may opt out of personalised advertising by visiting Google Ads Settings. Learn how Google uses information from sites that use its services. If you are in the European Economic Area, the United Kingdom or Switzerland, we ask for your consent before advertising cookies are used, and you can change your choice at any time using the button below.
Security Safeguards
Data is encrypted in transit (TLS) and at rest, and access is role-based and limited to staff who need it.
Account Deletion & Grace Period
When you request account deletion, your account enters a 30-day grace period, during which you can cancel by signing in. After 30 days, conversations with Kai, user records, personality metrics, compatibility data, and profile information are permanently purged from active systems, subject only to statutory financial retention requirements or pending safety investigations. We keep a minimal record that a phone number belonged to a deleted account, to prevent ban evasion.
Breaches
If a breach affects your personal data, we will tell you without undue delay, and we will notify the relevant regulator within 72 hours where the law requires it.
A. European Economic Area (EEA), United Kingdom, and Switzerland (GDPR / UK GDPR)
B. California Residents (CCPA / CPRA)
C. India (Digital Personal Data Protection Act, 2023)
Together with Kai is strictly restricted to individuals aged 18 and older. We do not knowingly solicit or collect data from minors. Any discovered minor accounts and associated records will be removed.
We reserve the right to amend this Privacy Policy periodically. For material changes, we will provide at least 30 days' advance notice via email or prominent in-app notification prior to implementation. Continued use of our platform constitutes acceptance of revised terms.
For inquiries, privacy requests, or formal notifications, contact our data protection team:
© 2026 Together with Kai. All rights reserved.