Content Monitoring Policy

1. Scope of monitoring

This policy covers the following surfaces of the Service:

• Uploaded posts and any media attached to them.
• Profile bios, profile media, and external links shared from a profile.
• Public comments and reactions.
• Direct messages, where direct messages are supported, are subject to pre-delivery automated moderation on shared media and to human review where required by law (for example, child-safety reports). Vyndle's messaging is not end-to-end encrypted by design, so that we can scan media at the platform level, action user reports, and respond to lawful authority requests.
• Links shared anywhere on the Service.
• Sponsored brand-brief content submitted by advertisers and the deliverables produced by creators in response.

2. Methods

We use a combination of methods to monitor the Service:

• Automated detection— every image and video uploaded to Vyndle is processed through the Microsoft Azure Content Safety API before it is made visible on the platform. Azure Content Safety classifies content across four harm categories — Hate and Fairness, Sexual, Violence, and Self-Harm — and returns a severity score for each. Vyndle applies graduated severity thresholds. Content scoring 5 or higher (on the Azure 0 to 7 scale) in any category is rejected before publication; it is not stored on the platform and is never shown to any user. Content scoring 3 or 4 is withheld from public view and flagged for review rather than published. Content scoring below 3 is published. These thresholds are set deliberately low, so that some benign content is withheld or reviewed in order to keep harmful content off the platform.
• User reports — anyone can flag content or accounts under our Reporting Policy.
• Human review — Vyndle staff or vetted moderation partners review flagged items, especially in high-severity or borderline cases.

Vyndle will also deploy voluntary perceptual hash-matching against the industry-standard known-CSAM hash lists maintained by the National Center for Missing and Exploited Children (NCMEC) and the Internet Watch Foundation (IWF), delivered through Cloudflare's CSAM Scanning Tool. This layer is not yet live: it requires enrolment with NCMEC and the IWF through Cloudflare and routing the media we serve through the Cloudflare proxy, and it will cover only content served that way. The automated detection described above operates today and does not depend on it. Section 20 of the DIS Industry Standard 2024 mandates proactive known-CSAM hash-matching only for Tier 1 services; Vyndle will implement it voluntarily as a defence-in-depth measure that exceeds the Tier 2 floor. Once it is live, any positive hash match will result in immediate blocking, evidence preservation, and mandatory reporting.

3. What we monitor for

We monitor for content and conduct that breaches our Terms or applicable law, including:

• The strictly prohibited categories in Section 3 of our Terms — content involving minors, graphic violence, politically sensitive propaganda or hate speech, sexual or pornographic material, and gambling promotion.
• Other illegal content under the laws of jurisdictions in which we operate.
• Intellectual-property infringement, including copyright and trademark abuse.
• Scam, phishing, and impersonation patterns.
• Abuse of platform mechanics — spam, fake engagement, payment fraud, chargeback abuse, and circumvention of moderation.

Vyndle classifies harmful content using the Australian regulatory framework. Class 1A covers child sexual exploitation material, pro-terror material that promotes, incites or instructs in terrorist acts, and extreme crime and violence material, as those terms are defined in the Online Safety (Designated Internet Services—Class 1A and Class 1B Material) Industry Standard 2024. Class 1B covers: (a) crime and violence material that depicts or advocates criminal activity or violence in a detailed or gratuitous way, but falls below the threshold of Class 1A extreme crime and violence material; and (b) drug-related material that promotes, instructs in, or facilitates the use, manufacture, or supply of controlled substances. Class 1Ccovers material that would be classified X 18+ — sexually explicit content and pornography. Class 2 covers material that would be classified R 18+ or MA 15+, including high-impact violence, detailed depictions or promotion of self-harm or suicide, and simulated gambling directed at minors. Class 1A and 1B material is subject to zero tolerance, mandatory reporting, and permanent account termination; Class 1C and Class 2 material is removed where it breaches our Terms.

4. Actions we may take

Where we identify a breach we may take any of the following actions, depending on severity and history:

• Take down the offending content.
• Apply an age-gate or visibility restriction to the content.
• Issue an account warning or temporary feature restriction.
• Suspend or terminate the account. Campaign payouts to it are then held by our payment provider rather than released; on a suspension, a held payout goes out on the next scheduled release run once the suspension ends. If the account is instead terminated, see our Creator Registration Policy for how its funds are handled.
• Escalate the matter to law enforcement where required.

Vyndle applies the following response-time targets to moderation actions:

• Class 1A and 1B(child sexual exploitation material, pro-terror material, extreme crime and violence material, and other Class 1B content) — immediate action; in any case within 24 hours of detection or report. We remove the content, preserve evidence, make a mandatory report to the Australian Centre to Counter Child Exploitation (ACCCE), the Australian Federal Police, and/or the eSafety Commissioner, and terminate the account.
• Class 1C and Class 2(age-restricted material out of policy) — within 24 hours of detection or report. We remove the content and warn or suspend the account depending on severity.
• Other Community Guideline breaches (harassment, hate, IP infringement, etc.) — within 72 hours of report. We remove the content and warn or suspend the account.
• DM-specific harms, where direct messages are supported (image-based abuse, grooming indicators, harassment in DMs) — within 24 hours of report, with immediate suspension of the conversation or account where indicators are severe.
• eSafety Commissioner removal notices— within the statutory period specified in the notice.

Detection of Class 1 material is reported to the Australian Centre to Counter Child Exploitation, the Australian Federal Police, and the eSafety Commissioner as required by Australian law. Account terminations for Class 1A or 1B material are final and not subject to appeal.

5. Appeals

If you believe a moderation decision was wrong, you can appeal in the app. When we remove your content we send you a notification that links straight to the appeal form, where you have 30 days from the decision to submit your appeal. The appeal form shows you the removed post — its content and dates — and the general categories it was reported under, never the identity of anyone who reported it. Include a brief description of why you believe the decision was incorrect and any supporting evidence. We aim to respond to appeals within 7 business days. Decisions on child-safety, sanctions, or law-enforcement matters are not subject to appeal.

6. Transparency

We may publish aggregate moderation statistics — for example, total takedowns by category over a period — to help users understand how the policy is applied. We will not publish details of individual cases or identify reported users in those reports.

7. Cooperation with law enforcement

Vyndle cooperates with valid legal requests from law-enforcement and regulatory agencies. Where we become aware of a credible, imminent threat to a person's life or to a child's safety, we may proactively report the matter and any associated evidence to the appropriate authorities, even without a formal request.

8. Retention of moderation evidence

Moderation logs — including automated moderation decisions, user reports, human-review actions, and account-level enforcement records — are retained for a minimum of 12 months, and longer where required by Australian law. Evidence relating to child sexual exploitation material is preserved under legal hold: it is exempt from routine deletion and kept under strict, access-controlled retention for as long as required, so that we can cooperate with the Australian Centre to Counter Child Exploitation, the Australian Federal Police, and the eSafety Commissioner. Retained evidence is used only for moderation, audit, and legal-response purposes.

9. Changes to this policy

We may update this policy from time to time. When we make material changes we will notify users by email or via an on-platform banner before the changes take effect.

10. Contact

General questions: hello@vyndle.co. To appeal a moderation decision, use the in-app appeal form linked from your removal notification (see Section 5).