Trust & Safety

Content Monitoring Policy

Last updated: September 8, 2026  ·  Effective immediately

Spark Twice is a curated, application-only matchmaking service. Every member is reviewed before they join, every profile is human-curated, and every message between members is sent through our platform. This policy explains how we monitor content across our service to keep the community safe, authentic, and respectful.

1. Scope

This policy applies to all content submitted to or transmitted through Spark Twice, including: applications, member profiles (photos, biographical text, values, preferences), member-to-member messages, support requests, and any other text or images shared on the platform.

2. Application review

Spark Twice operates by application only. No one creates an account by simply signing up. Every prospective member completes a multi-step application that includes identity details, date of birth, location, photos, and qualitative answers about themselves and the relationship they are seeking.

Each application is reviewed by our team before a membership decision is made. Applications inconsistent with the standards in our Community Guidelines are declined.

3. Automated content checks at application

Before an application reaches human review, the platform runs a series of automated checks to surface fraud and policy risks:

• IP and geolocation checks. Applications are scored against IP reputation data (VPN, proxy, Tor exit node, anonymous network, known threat IP) and geographic location signals.
• Disposable email detection. Applications from known throwaway email providers are flagged for review.
• Velocity checks. Repeated applications from the same IP within a short window are flagged.
• Behavioral signals. Honeypot field activation, abnormally short completion time, and other automation-like behavior are flagged.
• Reverse image search. Applicant photos are checked against the broader web using image search to detect reused, stolen, or stock imagery.

Flagged applications are not auto-rejected. They are routed to admin review with the specific signals attached so a human can decide.

4. Identity and liveness verification

Accepted members complete a live selfie verification before becoming active on the platform. The selfie is compared against the photos submitted during application using facial-recognition (biometric) technology. This processing, the consent it relies on, and how long the resulting face template is kept are described in our Privacy Policy. Members who fail liveness verification cannot send or receive messages and cannot be matched.

The platform also supports step-up re-verification: if any member is flagged for suspicious behavior after activation, a live selfie is required again before the account can continue. Repeated failures result in automatic suspension pending review.

5. Member-to-member message monitoring

All messages between members are sent through the Spark Twice platform. We do not facilitate off-platform contact. Our system performs automated content monitoring on every message to detect:

• Sharing of personal contact information. Email addresses, phone numbers, and common obfuscations (e.g. "at" in place of @, "dot" in place of .) are detected and flagged.
• Mentions of external platforms. Attempts to move conversations to other messaging or payment services (WhatsApp, Telegram, Signal, Snapchat, Venmo, CashApp, Zelle, and similar) are detected and flagged.
• Spam velocity. Messaging behavior that exceeds normal human use (high volume of messages across many conversations in a short window) is detected and flagged.

Flagged messages are not auto-deleted. They are added to an admin review queue with the specific signal attached so a human can decide whether to warn, restrict, or remove the member.

6. Profile content monitoring

Member profile updates (biographical text, spark statement, profession) are scanned with the same automated rules used on messages. Any profile text containing personal contact information or platform-redirection attempts is flagged for review.

Photos uploaded to a member profile are checked against the same reverse image search system used during application. Reused or known-stolen imagery is flagged.

7. Member reporting

Every member can report another member directly from a profile or a conversation thread. Reports include a free-text reason and route to our admin queue. We also provide every member with the ability to block another member from contacting them, immediately and without explanation.

Our 24-hour commitment. Spark Twice has zero tolerance for objectionable content and abusive behaviour. We review every report and act on reports of objectionable content or abusive users within 24 hours — which may include removing the content and ejecting the member who provided it. Reports that involve potential criminal activity may be shared with law enforcement.

8. Admin review and action

An administrator reviews flagged content, member reports, and the automated-signal queue. Actions available include: clearing a flag with no action; warning a member; temporarily pausing a member's profile pending further investigation; permanently removing a member from the platform; and removing or restricting individual profile content or photos.

All administrative actions are recorded in an internal audit log with timestamp, actor, target, and reason.

9. Prohibited content

The following content is prohibited on Spark Twice and will result in removal of the content and, depending on severity, removal of the member:

• Sexually explicit images of any kind
• Images depicting violence, weapons, or self-harm
• Hateful, discriminatory, threatening, or harassing language
• Solicitation of money, investment, gifts, or financial information
• Promotion of any external product, service, or platform
• Impersonation of another person, real or invented
• Photos of, or content involving, anyone other than the member themselves (with the exception of group photos clearly labelled)
• Any content that violates applicable law

10. Member education

Our Community Guidelines set out the conduct standards every member agrees to when joining. New members are encouraged to read the guidelines during onboarding, and the guidelines are linked from the dashboard, the messaging area, and the help section.

11. Policy updates

This policy may be updated to reflect changes in our service, our automated detection systems, or legal requirements. The "Last updated" date at the top of this page reflects the most recent revision.