Documentation / Legal / Discord>Discord Bot Privacy Policy

Discord Bot Privacy Policy

•Updated November 20, 2026

Our privacy commitments, zero chat storage policy, and automated data retention hygiene for Arqora's Discord bot.

01.Built Exclusively for Arqora

The Arqora Discord bot is a private, custom-engineered security system built strictly for the official Arqora Discord community (supporting arqora.be and arqora.work).

It is not a public bot. It cannot be invited to outside servers, it does not track user activity across third-party communities, and it does not aggregate cross-server profiles. Its sole job is to protect our own developers, contributors, and community members from raids, phishing, account hijackings, and malicious spam.

We believe privacy policies should be straightforward, honest human commitments rather than dense corporate boilerplate. We never sell personal data, we never run targeted advertising, and we never archive your conversations.

02.Zero Chat Storage (RAM-Only Inspection)

The most important privacy guarantee we make is simple: We do not store your chat messages.

  • Volatile Memory Processing: When you send a message in the server, the bot inspects the content purely in volatile RAM for milliseconds to scan for high-risk threats: known phishing domains, deceptive punycode links, unauthorized broadcast mentions (@everyone / @here), and accidentally leaked credentials (such as Discord tokens, GitHub personal access tokens, AWS keys, or private keys).
  • Immediate Discard: Once the security check runs, the text is immediately dumped from memory. Nothing is written to disk, databases, or third-party loggers.
  • Violation Metadata Only: If a message violates server security rules (such as a malicious phishing link), only the action metadata (user ID, timestamp, the specific blocked domain, and rule triggered) is retained for staff audit logs. The surrounding conversation is never saved.

03.What We Actually Store (Minimal Technical State)

We store only the absolute bare minimum data needed to protect the server and keep things running smoothly:

  • Discord User Identifiers: Your Discord User ID, username, and account creation timestamp. We use this to calculate basic join safety scores (e.g., catching brand-new throwaway accounts created 5 minutes ago) and to remember whether your account has completed human verification.
  • Verification Tokens: If an account is placed in security quarantine, a temporary single-use UUID token is generated. This token is used strictly to link your Discord session to the verification portal and automatically expires after 24 hours.
  • Security & Strike Records: To avoid punishing honest mistakes, we track active warnings and link infractions across a 14-day sliding window. This powers our graduated strike ladder (Strike 1: friendly notice, Strike 2: 5-minute timeout, Strike 3: 1-hour timeout, Strike 4+: quarantine review).
  • Cryptographic Audit Trail: Staff moderation actions (warnings, timeouts, quarantine resolutions) are recorded in an internal SQLite audit log. To prevent tampering, rogue staff abuse, or history modification, each audit entry is cryptographically linked with SHA-256 hashes.

04.Automated Data Retention & Hygiene Lifecycle

We follow strict data minimization principles and Discord's Developer Policy. Our bot runs an automated background cleanup worker every 6 hours that prunes old database records according to our scheduled retention lifecycle:

24 Hours
Verification tokens expire and become invalid.
48 Hours
Expired, unredeemed tokens are permanently purged from disk.
7 Days
Successfully redeemed tokens are wiped from the database.
90 Days
Inactive member profiles with clear status are removed.
180 Days
Inactive verified members and expired warnings are wiped clean.
365 Days
Historic security audit entries are permanently archived and deleted.

Automatic Risk Decay: Accounts that maintain a clean record have their risk scores automatically decayed by 10 points for every 30 days of good standing. We do not maintain permanent grudge scores.

05.Self-Service Privacy & Data Deletion (/privacy)

You should never have to wonder what a bot knows about you. We built native self-service privacy commands directly into the bot:

  • /privacy view: Outputs a private, ephemeral card showing your exact stored profile: your join timestamp, verification status, active risk score, recorded infractions, and pending tokens.
  • /privacy delete: Allows you to exercise your GDPR right to erasure instantly. Running this command deletes your member record, warning history, and verification tokens from our SQLite database on the spot.
  • Disciplinary Investigation Hold: The only temporary exception to instant deletion is if an account is currently under an active staff security investigation (such as coordinated server raids or malicious credential phishing). This hold prevents bad actors from evading bans by cycling deletions. Once staff resolves the investigation, standard deletion rights apply.
  • Email Requests: You can also email [email protected] at any time for manual data deletion or inquiries.

06.Web Verification & Cloudflare Turnstile

When quarantined accounts open their private link on verify.arqora.be:

  • Privacy-Preserving Turnstile: We use Cloudflare Turnstile to confirm that the user is human. Turnstile does not use cross-site tracking cookies, does not build advertising profiles, and never sells data.
  • No IP Storage: Your IP address is evaluated strictly in-flight by Cloudflare to prevent botnet floods. Our bot server never logs or databases your IP address.
  • Instant Role Restoration: Once completed, the bot automatically removes the @Infected role on Discord, marks your account verified, and cleans up the quarantine notice in #infected.

07.Quarantine Isolation & Recidivist Disciplinary Hold

To keep the community safe while avoiding endless verification loops, the bot enforces clear boundary isolation:

  • Total Channel Isolation: When an account is placed under a security hold, our automated permission engine ensures that all server channels and categories deny access (ViewChannel: false, SendMessages: false). Quarantined members only see the designated #infected channel.
  • Recidivist Hold (3/3 Strikes): We do not trap members in endless captcha loops. If an account repeatedly violates link or spam policies and reaches 3 active strikes, self-service captcha verification is disabled. The account is placed on a formal Disciplinary Hold requiring human staff review.
  • Right to Appeal: Anyone placed under a disciplinary hold can submit a formal appeal using the [📝 Submit Appeal] button in #infected or by running /appeal. Appeals are reviewed by human server moderators within 24 to 48 hours.

08.Contact & Privacy Inquiries

If you have questions, feedback, or need help with your privacy settings or moderation history, please reach out directly:

Privacy & Legal: [email protected] or [email protected]

General Inquiries: [email protected]

These policies may be updated as Arqora evolves. For legal inquiries, please contact [email protected].

ArqoraDiscord Bot Privacy Policy
Bot privacy policy active — November 20, 2026