Content Moderation Policy

Plum | Effective Date: September 29, 2026 | Version 1.0

1 Prohibited Content

Plum is an AI-powered interactive entertainment service operated by Within Tech Limited (“we,” “us,” or “our”). This Content Moderation Policy supplements our Terms of Service and Community Guidelines and explains our content restrictions, moderation practices, reporting channels, and response commitments.

This Policy applies to content created, uploaded, requested, generated, published, or shared through Plum, including character profiles, images, user inputs, AI-generated responses, chats, and memories. It applies to both public content and private interactions.

Moderation Mechanisms

We use automated content checks and human review to identify and address prohibited content. Character text and images submitted for publication are subject to moderation. Content may be rejected, returned for changes, or referred for human review. We may also review published content and reported interactions, block generation or publication, restrict visibility, remove content, and take appropriate account action.

For suspected deepfake or synthetic-media abuse, we assess the content, its context, reports received, and relevant authorization records. Our review considers whether a real person is identifiable, whether appropriate consent or authorization exists, whether the content is presented as authentic, and whether it involves sexual exploitation, impersonation, harassment, fraud, or other harm. We may request supporting information and restrict access to content while reviewing it.

Users must not bypass safety controls or re-upload content to evade enforcement. Automated checks and human review may make mistakes and cannot identify every violation; reporting and appeals provide additional safeguards.

The following categories of content are prohibited on Plum. Confirmed violations will result in content removal or restriction and appropriate account action, up to and including permanent termination of access. Urgent protective measures may be taken before an investigation is complete.

Deepfake — Permitted vs. Prohibited

Permitted, subject to applicable law and all other Plum policies:

• Clearly labeled artistic or satirical content disclosed as AI-generated and unlikely to deceive, provided it respects the depicted person’s rights.
• Use of a real person’s likeness with valid written authorization covering the intended use. Users must retain supporting records and provide them to Plum for verification upon request.
• Fictional characters with no identifiable real-world counterpart.

Prohibited:

• Placing real people in fabricated scenarios presented as factual.
• Non-consensual intimate or sexual depictions.
• Unauthorized commercial use of a real person’s likeness, identity, or voice.
• Content designed to harass, defraud, exploit, or intimidate a specific individual.

Labeling content as fictional, satirical, or AI-generated does not make otherwise prohibited content permissible. Authorization to use a person’s likeness does not automatically make content compliant with this Policy. All uses remain subject to the applicable content restrictions.

2 Reporting Mechanism

2.1 Report Button

Plum provides a clearly labeled “Report” or “Report Content” entry point on content display pages so that users can report content directly without first contacting customer support.

Placement and accessibility:

• Available on content cards, character detail pages, image viewers, and other applicable content display surfaces.
• Accessible through an overflow menu or a standalone reporting control.
• Clearly labeled “Report” or “Report Content.”
• Available without login for publicly accessible content.
• Equivalent reporting access will be provided in any Plum iOS or Android application made available to users.

Reports may concern content, an account, or a specific interaction. Please include the relevant link or identifier, the reason for the report, and only the information necessary to investigate.

Do not download, copy, or redistribute suspected CSAM to submit a report. Provide its location or identifier instead. Do not include passwords, payment credentials, or unrelated sensitive information.

2.2 Dedicated Reporting Contacts

In addition to the in-platform Report button, the following dedicated email addresses are checked seven days a week.

You do not need a Plum account to report by email. Reports are prioritized by their substance and risk, regardless of which designated channel receives them.

If someone faces immediate danger, contact local emergency services as well. Plum is not an emergency response service.

2.3 Reportable Violation Categories

Users may report the following through the Report button or the designated email channels:

• Deepfake / Synthetic Media — a real person depicted without consent or used deceptively.
• Non-Consensual Intimate Imagery (NCII).
• Child Safety (CSAM) — sexual content or exploitation involving minors.
• Hate Speech / Discriminatory Content.
• Violence / Threatening Content.
• Harassment / Impersonation.
• Misinformation / Fabricated Content.
• Copyright / IP Infringement.
• Spam / Scam.
• Other Policy Violation.

2.4 Resolution Commitments (SLA)

The following timeframes run from receipt of a report or appeal. “Hours” means elapsed hours, including weekends and public holidays. “Business days” means Monday through Friday, excluding public holidays in Hong Kong.

Response Commitment

Plum will conduct an initial human assessment of urgent safety reports and provide a response within 48 hours of receipt. For non-urgent reports, Plum will acknowledge receipt within 5 business days and communicate a decision or substantive investigation update within 15 business days. Where further investigation is required, Plum will explain the status, provide an estimated completion timeline, and send further updates at least every 15 business days until closure.

These are response and review commitments, not a guarantee that every investigation will be concluded within the initial response period. An ongoing investigation does not delay necessary protective action. Any shorter timeframe required by applicable law takes precedence.

Individual acknowledgements and updates can only be sent where the reporter provides a usable contact method.

2.5 Post-Report Process

• Acknowledgement — Reporters who provide a usable contact method receive confirmation of receipt within the applicable timeframe. Subsequent communications may refer to a report reference number or the original email.
• Review — Designated personnel assess the report against this Policy and our Community Guidelines, supported by automated checks where appropriate. We may request additional information necessary to investigate.
• Action — Confirmed violations may result in content removal, restricted visibility, generation or publication restrictions, an account warning, suspension, or permanent termination. Necessary protective measures may be taken before the investigation is complete.
• Notification — The reporter receives a decision or investigation update within the applicable timeframe. Disclosures may be limited to protect privacy, safety, or an investigation.
• Appeal — A reporter or affected user may dispute a decision within 30 calendar days of notification by replying to the notification or contacting a designated reporting email address. Please include the original email or report reference, the reasons for the appeal, and relevant supporting information. A designated review lead reviews the appeal and provides a decision or progress update within 15 business days. An appeal does not automatically suspend the original action, and review does not guarantee a different outcome.

2.6 Reporter Protections

Good Faith — Users will not face adverse action merely for submitting a good-faith report, even if the report does not result in content removal.

Confidentiality — We will not disclose a reporter’s identity to the subject of a report without consent, except where required by law.

No Retaliation — Plum prohibits retaliation against anyone for submitting a good-faith report.

Anonymous Submissions — Reports may be submitted without providing a name or creating a Plum account. Providing a contact method enables us to send acknowledgements and resolution updates. Anonymous reporting does not mean that no technical information is processed; information handling is described in our Privacy Policy.

Quick Reference — Reporting Contacts