ChatSe Alpha — AI Journal
Plain summary. ChatSe is an early-stage product. We send your messages to AI providers to answer them. We store what you tell us so you can find it later. You can delete your data anytime by messaging “forget me” (Section 7 says exactly what that covers today). We will email or message you if we change these terms in a material way. We don’t sell your data. We don’t guarantee uptime. Use it as a notebook, not as a system of record for medical, legal, or financial decisions.
These Terms govern your access to ChatSe Alpha, operated by TCL. By interacting with ChatSe over WhatsApp, iMessage, email or our web sign-up (sending a message, replying “I agree”, forwarding a link or an email, signing up on the web, or accepting a shared-journal invite) you agree to these Terms.
If you do not agree, stop messaging the service. To erase what we already have, send “forget me” — see Section 7.
ChatSe Alpha is a chat-based personal assistant that:
ChatSe is NOT:
You must be at least 13 years old to use ChatSe (16 in jurisdictions that require it, e.g. EU GDPR-K). You may not use ChatSe on behalf of someone else without their explicit consent.
We operate ChatSe in the United States. Using it from outside the US is at your discretion and risk.
You may:
You may not:
We may suspend or remove your access (and may erase your data) if you violate these rules.
Because ChatSe is conversational, almost everything you send is data we process. We try to be specific about the categories:
| Category | Examples | Why we keep it |
|---|---|---|
| Messages | What you type / dictate / forward | Required to respond + searchable retrieval |
| Media | Photos, voice notes, documents, vCards | We extract structured info; raw bytes stored in private GCS for 90 days |
| Schedule data | Todos, reminders, briefing times, timezone | Required for the feature to work |
| Journal data | Journals you create, members, items posted | Required for sharing |
| Memories | Long-lived facts you ask us to remember | Required for personalization across sessions |
| Google data (optional) | Calendar events, Gmail metadata + bodies | Only when you connect Google; read on a schedule for your briefing and event alerts, and when you ask; processed by the AI providers in Section 6 |
| Web sign-up | A waitlist email; the phone number and email you verify; the name and avatar you pick | To invite you and set up your account; a waitlist email is deleted after 6 months |
| System logs | Phone number, message ids, timestamps, latency, costs, and today some message text | Operational + abuse / billing reconciliation; kept up to 30 days |
| Conversation records | The prompts and replies of each AI call, including your messages | Fixing problems and improving the service; kept until deleted (see Section 7) |
Sensitive categories. Medical and financial information is stored like everything else today: in our private database and storage, encrypted at rest by Google Cloud, but not separately encrypted by us. Don’t send passwords, PINs or full card numbers.
What we do NOT collect:
gmail.readonly).We share the minimum necessary to deliver the service:
We do not sell your data. We do not share it with advertisers.
If a court or law-enforcement request compels disclosure, we will limit it to what’s strictly required and notify you when we are legally allowed to.
You can, at any time, by messaging ChatSe:
To disconnect Google at any time, remove GIRI at myaccount.google.com/permissions. That revokes our access to your Calendar and Gmail.
For full data export, email giri@thecontextlab.ai with the phone number or email you use with ChatSe. We’ll deliver a JSON archive within 30 days.
If you have a question about a specific data point (correction, clarification), email the same address.
When you join or create a shared journal:
When you accept a shared-journal invite, you accept that journal’s own short-form ToS (a one-line “by joining you agree the creator can share items into this journal with you and other members”).
ChatSe’s responses come from large language models. They are sometimes wrong. You agree to:
When ChatSe forwards your content to Anthropic, Google, or OpenAI, you warrant that the content does not violate those providers’ usage policies (including no CSAM, no instructions for weapons of mass destruction, no content that would force the model into deceptive behavior). If we detect a violation, we may refuse the request, notify you, and in serious cases suspend access.
THE SERVICE IS PROVIDED “AS IS” AND “AS AVAILABLE.” TO THE MAXIMUM EXTENT PERMITTED BY APPLICABLE LAW:
These limits do not exclude liability that cannot be excluded under applicable law (e.g. fraud, gross negligence, statutory consumer rights).
You agree to indemnify and hold harmless TCL, its officers, and employees from any claim arising out of (a) content you submit, (b) your violation of these Terms, or (c) your violation of a third party’s rights. We will give you reasonable notice and let you control the defense (using counsel reasonably acceptable to us).
You can stop using ChatSe at any time and erase your data with “forget me” (Section 7).
We may terminate or suspend your access if you violate these Terms or if we sunset the alpha. We will give 30 days’ notice for a planned sunset and provide an export window.
We may update these Terms. For material changes (anything that expands what we share, narrows your rights, or changes data retention) we will give at least 14 days’ notice by a message on the channel you use with ChatSe and update the Effective date at the top. Continued use after the effective date constitutes acceptance.
These Terms are governed by the laws of the State of California, USA, without regard to conflict-of-law rules. Disputes must be filed in the state or federal courts located in San Francisco County, California. You waive any objection to that venue. EU residents retain mandatory protections under their local law.
Questions, takedown requests, data exports, security reports: giri@thecontextlab.ai
For security reports (auth, data exposure), email the same address with “security” in the subject line.
| Version | Date | Notes |
|---|---|---|
| v1 | 2026-05-04 | Initial alpha ToS — covers WhatsApp transport (Twilio + future Meta), AI providers, shared journals, sensitive-data sealing, /forget self-delete, 90-day media retention. |
| v1.1 | 2026-10-03 | Corrections after an internal privacy audit: what “forget me” clears today, how sensitive data is stored (no separate sealing), the full provider list (OpenAI voice, Sendblue, SendGrid/Twilio Verify, Resend inbound), log contents and retention, conversation records. No new uses of your data. |
| v1.2 | 2026-10-04 | “forget me” now also clears the search index, conversation records and audit entries about you; system logs mask phone numbers and message text. |
| v1.3 | 2026-10-13 | Contact address is now giri@thecontextlab.ai (the old address still reaches us). No other change. |
| v1.4 | 2026-10-13 | Correction: Section 7 listed a “disconnect google” message command that ChatSe does not have. It is removed; to disconnect Google, remove GIRI at myaccount.google.com/permissions. No change to how we use your data. |
| v1.5 | 2026-10-13 | Corrections so the text matches the service: ChatSe works over WhatsApp, iMessage, email and the web, and its assistant is called GIRI; Google data is read-only and read on a schedule for your briefing and alerts, and when you ask; the provider list now names Cloudflare (it relays incoming messages) and Gemini’s web search and YouTube use; web sign-up data is listed; security reports go by email. No new uses of your data. |
This document is the canonical text. The hosted copy at https://www.thecontextlab.ai/ChatSe-ToS may be reformatted but must not change wording without bumping the version above.