Trust & Compliance

How Connections handles your data — engineered to industry-standard security controls (SOC 2 / ISO 27001 / NIST 800-53 / GDPR / DPDPA aligned) with end-to-end encrypted messaging and tamper-evident audit logs.

Last reviewed: April 2026

How we protect your data

End-to-end encryption

Direct messages are encrypted using the Signal Protocol (X3DH key agreement + Double Ratchet) via libsodium. Only sender and recipient can read message content — Connections servers never see plaintext.

Encryption at rest & in transit

All client-server traffic uses TLS 1.2+. Local message storage is encrypted with SQLCipher (AES-256). Server-side data uses MongoDB encryption at rest and AWS S3 server-side encryption.

Tamper-evident audit logs

Every privileged admin action is recorded in a SHA-256 hash-chained audit log. Modifying past entries breaks the chain — detectable on every verification cycle.

Immutable 7-year archive

Audit records sealed daily to AWS S3 Object Lock in COMPLIANCE mode for 7 years. Even root AWS access cannot delete sealed records before retention expiry.

Continuous monitoring

Real-time alerting on suspicious admin patterns (burst deletions, failed-login bursts, PII fishing, new-IP logins, chain integrity breaks). Critical events page on-call within minutes.

Privacy by design

GDPR Art. 25 honored: passwords, tokens, and payment fields are redacted at write-time before reaching any log. Region-aware consent management (GDPR / CCPA / DPDPA / LGPD / PIPEDA / APPI).

Strict access controls

Admin panel uses Google / Facebook OAuth with MongoDB allowlist. Four-tier RBAC (superadmin / auditor / admin / viewer). Every admin action is logged and reviewable.

Automated content safety

Image uploads pass through NudeNet (NSFW), age estimation ML, and AI text moderation (Google Gemini / OpenAI) before reaching public surfaces. NSFW detection in chat runs on-device so server never sees decrypted media.

Compliance posture

We engineer to industry-standard control frameworks. Each row below describes the framework, what it requires of audit-and-access controls, and our current alignment.

Framework Scope Status Evidence
SOC 2 Type II Common Criteria CC4 (audit), CC6 (access), CC7 (monitoring), CC8 (change) Controls aligned Hash-chained audit log, RBAC split, alerting cron, before/after diffs on mutations
NIST 800-53 r5 AU family (Audit & Accountability) — AU-2 through AU-12 Met Centralised emitter, cryptographic non-repudiation, content + retention + protection
ISO 27001:2022 A.8.15 Logging, A.8.16 Monitoring, A.5.34 PII, A.5.10 Acceptable use Controls aligned Logging policy enforced; PII-read allowlist; access reviews via admin RBAC
GDPR (EU) Art. 5 (integrity), 25 (privacy by design), 30 (records of processing), 32 (security) Met Region-aware consent; DSR endpoints; PII redaction; 7-year processing records
CCPA / CPRA (California) §1798.100 right to know; §1798.105 right to delete Met Data export endpoint; account-deletion flow; consent toggle
DPDPA 2023 (India) §7 consent; §8 obligations; §11 right to correction/erasure Met Granular consent; localized DSR; retention policy 1-3y aligned
PCI DSS 4.0 §10 (Logging) — applies to payment-webhook surface Controls aligned 10.2 content / 10.5 protection / 10.7 retention all met
App Store / Google Play Audit trail for moderation + financial dispute history Met All moderation decisions logged with diff + reason; 7y subscription history
SOC 2 Type II Certification External audit firm sign-off Roadmap Controls in place; engagement with auditor planned 2026 H2

Status legend: Met — all listed controls implemented and verifiable. Controls aligned — controls implemented; pending external audit attestation. Roadmap — planned within next 12 months.

Key technical controls in place today

Hash-chained audit logs

Every admin action gets a sequence number, prevHash, and SHA-256 entryHash. Daily integrity verification cron alerts on any break.

WORM 7-year archive

S3 Object Lock COMPLIANCE mode with 2557-day retention. Immutable even to root AWS account.

SIEM streaming

Audit events teed to dedicated Kafka topic for downstream Splunk / Datadog / ELK consumption.

Field-level change tracking

Every state-changing admin action records before-and-after values for affected fields.

Suspicious-pattern alerting

5 detection rules running every 60s: burst-delete, burst-block, failed-login bursts, new-IP logins, PII fishing.

RBAC with separation of duties

Auditor role can read full audit trail but cannot modify user data. Plain admins see only their own actions.

NoSQL injection protection

Mongoose security plugin blocks $where, $expr, dangerous operators globally. HTTP parameter pollution + XSS guard at request layer.

HMAC request signing

Every client request signed with HMAC-SHA256. Replay protection via 2-min timestamp window + Redis nonce store.

Rate limiting + fraud velocity

Redis-backed cluster-aware rate limiter on auth endpoints. Per-user fraud signals (signup velocity, like-rate, message-rate).

Region-aware consent

13 strict-opt-in regulatory regions detected automatically (GDPR, DPDPA, LGPD, PIPEDA, APPI, PIPA, APA, POPIA, VCDPA, CPA, CTDPA, UCPA, CCPA).

Strict 180-day hot retention

Audit log MongoDB collection auto-prunes via TTL; sealed copy preserved 7y in WORM archive.

Distributed-trace correlation

OpenTelemetry traceId on every audit row — joins audit event to any downstream service trace.

Data processing & subprocessors

We use the following providers to deliver Connections. Each subprocessor has been evaluated for security, signed appropriate data-processing agreements, and (where applicable) maintains its own SOC 2 / ISO 27001 attestations.

Subprocessor Purpose Region Their attestations
Amazon Web Services (AWS)Compute, storage (S3), CDN, Object Lock archiveap-south-1 (primary), us-east-1 (CDN)SOC 2 Type II, ISO 27001, PCI DSS
Google FirebasePush notifications (FCM)us-central1SOC 2 Type II, ISO 27001
Google Gemini APIAI bio generation, content moderationus-central1SOC 2 Type II, ISO 27001, ISO 27018
OpenAI API (fallback)AI content (when Gemini unavailable)USSOC 2 Type II
Stripe / Apple / Google PlayPayment processingUS / EU / INPCI DSS Level 1, SOC 2 Type II
Nexmo (Vonage)SMS OTP deliveryGlobalISO 27001

Incident response

We maintain an incident response plan with defined severity levels, escalation paths, and notification timelines. In the event of a confirmed personal data breach:

  • EU users (GDPR Art. 33): notified to relevant Data Protection Authority within 72 hours of confirmation.
  • California users (CCPA): notified per §1798.82 timing requirements.
  • India users (DPDPA): notified to Data Protection Board per §8(6) requirements.
  • Affected users: direct notification within 72 hours unless law enforcement requires delay for an active investigation.

Suspected security issues should be reported to connections@dating-universe.com with subject line "Security Vulnerability Report". See our Security Policy for full responsible-disclosure guidelines.

Documents & reports

For enterprise customers, procurement teams, or compliance reviews — request a security questionnaire, sample audit-log report, or our written policies via the contact below. We respond within 5 business days.

For compliance questions: connections@dating-universe.com
For security disclosures: connections@dating-universe.com (subject "Security Vulnerability Report")