Content Monitoring & Trust and Safety Policy

1. Purpose and Scope

This Content Monitoring Policy ("Policy") establishes the mandatory standards, workflows, and controls that PLADA TECHNOLOGIES LTD ("the Company") applies to every piece of user-generated or licensed content uploaded to, streamed on, or hosted by its platform(s). The Policy applies to all employees, contractors, third-party moderation vendors, automated systems, and platform partners involved in the ingestion, review, storage, or distribution of content.

The Policy exists to:

• Ensure zero tolerance for child sexual abuse material (CSAM), non-consensual content, and other categories of illegal material.
• Verify that every performer depicted in adult content is a documented, consenting adult.
• Maintain compliance with U.S. federal record-keeping law (18 U.S.C. §§ 2257/2257A), CSAM reporting obligations (18 U.S.C. § 2258A), FOSTA-SESTA, and applicable state and international regulations (e.g., UK Online Safety Act, EU Digital Services Act).
• Provide a consistent, auditable, and humane review process for both automated systems and human moderators.

This Policy applies regardless of monetization status, where all of the content is created by artificial-intelligence-generated models.

2. Definitions

• CSAM: Child Sexual Abuse Material, including any visual depiction of a minor (under 18) engaged in sexually explicit conduct, or content that is digitally generated, AI-synthesized, or altered (e.g., deepfake, 'de-aged') to depict a minor in such conduct.
• Non-Consensual Content (NCC): Material depicting a real person in a sexual or intimate context distributed without that person's knowledge or consent, including 'revenge porn', hidden-camera/voyeur content, and non-consensual intimate imagery generated or altered by AI (synthetic/deepfake NCC).
• Age and Identity Verification (AIV): The process of confirming, through government-issued identification and liveness matching, that every performer is at least 18 years old and is the person depicted.
• Trusted Flagger: A vetted individual or organization (e.g., NCMEC, INHOPE member hotline, law-enforcement liaison) whose reports receive expedited review.
• Escalation Tier: The internal severity classification (Tiers 1-3, see Section 8) that determines review urgency and required sign-off.

3. Governing Principles

• Zero tolerance, always: Any confirmed or suspected CSAM is never allowed to be generated by the AI models. No exception, appeal, or business justification overrides this rule.
• Verify before publish: No content is published without review, and all content is generated by artificial-intelligence models.
• Defense in depth: No single control (automated or human) is treated as sufficient; every category of prohibited content is checked by at least two independent mechanisms.
• Continuous improvement: Detection models, keyword lists, and reviewer guidance are updated on a recurring basis in response to emerging harm patterns (e.g., new AI-generation techniques).

4. Prohibited Content Categories

The following categories are prohibited in all circumstances and trigger immediate removal. Categories marked (*) additionally trigger mandatory external reporting per Section 9.

4.1 Absolute Prohibitions

• Child sexual abuse material (*): any real, animated, or AI-generated depiction of a minor in sexually explicit conduct, including content that stylizes or captions adult performers as minors ('age-play' marketed as underage).
• Non-consensual content (*): hidden-camera footage, distribution without subject consent, and non-consensual deepfakes/synthetic imagery of real people.
• Incest-themed content presented as factual/real, bestiality, necrophilia, and depictions of non-simulated violence, mutilation, or death are not allowed to be generated by the AI models.
• Content depicting or soliciting acts without documented consent from all depicted parties, including 'cheating'/spycam AI content, is never allowed to be presented on the site.

4.2 Restricted / Requires Additional Controls

• Fetish content involving simulated age regression, or costuming/props creating ambiguity about a performer's age (rejected pending manual age-plausibility review) by the AI models.
• AI-generated or AI-modified sexual content of any kind, held pending provenance and consent verification per Section 6.3.

5. Age and Identity Verification (Onboarding Controls)

No individual may view content until the following are completed and stored per Section 13:

• Government-issued photo ID confirming the performer is 18 years of age or older.
• A signed model release / consent form for each specific piece of content, referencing the content by unique production ID.
• A real-time liveness check (selfie-with-ID or video attestation) matching the performer to the submitted ID.
• Secondary ID/database verification (e.g., government ID validation service) for jurisdictions with elevated fraud risk.

Records required under 18 U.S.C. § 2257/2257A are compiled into a Custodian of Records file indexed by ID, and are made available for regulatory inspection per statutory requirements.

6. Content Upload and Pre-Publication Review Workflow

6.1 Intake

Every file, on upload, is automatically hashed, metadata-scanned, and routed into a review queue before it becomes visible to any user other than the uploader.

6.2 Automated Screening (see Section 7 for tooling detail)

Machine-learning classifiers score for coercion indicators, weapons, and other prohibited categories.

6.3 AI-Generated Content

Content that is wholly AI-generated or AI-modified is withheld from publication until it passes the same CSAM and NCC screening as any other upload. Synthetic content that could plausibly depict a real, identifiable minor is treated as CSAM under Section 4.1 regardless of claimed AI origin.

7. Automated Detection Technology

• Hash-matching databases: Integration with recognized CSAM hash-sharing programs (e.g., NCMEC's hash database, industry hash-sharing consortia) to block known illegal files at upload, before human eyes see them.
• Perceptual/AI classifiers: Machine-learning models score uploads and only allow AI-generated content to be uploaded on the site.
• Text and metadata scanning: Automated keyword and caption scanning for age-related, coercion-related, or trafficking-related language across titles, tags, captions, and comments.
• Model governance: All detection models are version-controlled, tested for false-negative rate on CSAM/NCC classes before deployment, and never used as the sole basis for publishing content, only for blocking or escalating it.

8. Human Moderation Review Process

Content and reports that require human judgment are triaged into escalation tiers:

• Tier 1, Critical (review within 1 hour, 24/7 coverage): suspected CSAM, suspected trafficking/coercion, imminent-harm reports, law-enforcement requests.
• Tier 2, High (review within 24 hours): suspected non-consensual content, suspected underage performer without confirmed evidence, repeat-offender accounts.
• Tier 3, Standard (review within 5 business days): policy/tagging disputes, quality and categorization issues, routine user reports.

Tier 1 determinations require sign-off from a senior moderator or Trust & Safety lead before final action and reporting. All decisions are logged with reviewer ID, timestamp, evidence reviewed, and rationale.

9. Mandatory External Reporting

Where U.S. law applies, the Company complies with 18 U.S.C. § 2258A. Upon identifying apparent CSAM, the Company:

• Immediately restricts public access to the material and preserves it in a secure evidence store.
• Files a report with the National Center for Missing & Exploited Children (NCMEC) CyberTipline as soon as reasonably possible, including all legally required content and account information.
• Preserves the reported material for the statutory retention period (not less than 90 days, or longer if requested by NCMEC or law enforcement) and does not further view, copy, or distribute it beyond what reporting requires.
• Cooperates with subsequent law-enforcement legal process (preservation letters, subpoenas, search warrants) through designated legal counsel.

For non-U.S. jurisdictions, the Company reports to the applicable national hotline (e.g., INHOPE network members) and law enforcement as locally required. Suspected trafficking or coercion is separately referred to law enforcement and, where applicable, national human-trafficking hotlines.

10. User and Trusted-Flagger Reporting Mechanisms

• A persistent 'Report' control is available on every piece of content and every user profile, requiring no account login to submit a report of suspected CSAM or non-consensual content.
• Reports alleging CSAM or non-consent are auto-routed to Tier 1 queues regardless of reporter-selected category.
• Trusted Flaggers (NCMEC, INHOPE hotlines, verified law-enforcement liaisons) have a dedicated, expedited submission channel and point of contact.
• Reporters receive a case reference number; where legally permissible, they are notified of outcome without disclosing confidential moderation detail.

11. Takedown, DMCA, and Right-of-Removal Requests

• Copyright (DMCA) takedown notices are processed by the designated DMCA agent within statutory timeframes, with counter-notice procedures available to uploaders.
• Any depicted individual may request removal of content showing them, without needing to prove non-consent, pending verification of identity; such requests are treated as Tier 2 minimum.
• Confirmed non-consensual content and CSAM are removed permanently and are not restored under any counter-notice or appeal.

12. Moderator Training, Support, and Wellness

• All moderators complete mandatory training covering legal definitions of CSAM and NCC, red-flag indicators of coercion/trafficking, escalation procedures, and evidentiary handling requirements, prior to reviewing live content.
• Ongoing refresher training occurs at least quarterly and whenever policy or detection tooling materially changes.
• Mandatory wellness measures are in place for reviewers exposed to graphic or distressing material, including exposure limits, mandatory breaks, rotation between content categories, and access to confidential mental-health support/counseling.
• Reviewer performance is audited via quality-assurance sampling with calibration sessions to ensure consistent application of this Policy.

13. Record-Keeping and Retention

• 2257 records (ID, model release, liveness verification) are retained for the full period required by law and stored in an access-controlled Custodian of Records system, separate from general content storage.
• Moderation decision logs (reviewer, timestamp, evidence, rationale, escalation path) are retained for a minimum of 3 years or longer where required by law or active legal process.
• CSAM/NCC reports and associated evidence are retained per the requesting agency's instructions (NCMEC/law enforcement), stored encrypted, and access-limited to designated legal/compliance personnel only.
• Access to retained sensitive records is logged, reviewed periodically, and restricted on a least-privilege basis.

14. Appeals Process

• Users and performers may appeal a content removal or account action, except for confirmed CSAM or confirmed non-consensual content determinations, which are final and not subject to appeal or reinstatement.
• Appeals are reviewed by a moderator not involved in the original decision and are resolved within 5 business days.
• Appeal outcomes and rationale are logged in the same audit system as the original decision.

15. Third-Party and Platform Partner Compliance

• Studio and affiliate content partners must contractually warrant complete AIV documentation for every performer and grant the Company audit rights over their records.
• Payment processors, CDNs, and hosting partners are selected and reviewed for their own CSAM-detection and reporting compliance; partner agreements require immediate mutual notification of any confirmed illegal content.
• The Company reserves the right to immediately suspend any partner feed found to contain CSAM, non-consensual content, or falsified AIV documentation, pending investigation.

16. Incident Response Plan

• A designated Trust & Safety on-call rotation provides 24/7 coverage for Tier 1 escalations, including out-of-hours law-enforcement contact.
• Confirmed CSAM or trafficking incidents trigger an incident-response protocol involving Legal, Trust & Safety leadership, and, where required, external counsel and law-enforcement liaison, within defined internal SLAs.
• Post-incident reviews are conducted for all Tier 1 events to identify detection or process gaps and drive tooling/training updates.

17. Governance, Oversight, and Audit

• A named executive (e.g., Chief Trust & Safety Officer or equivalent) owns this Policy and reports compliance metrics to senior leadership and, where applicable, the Board on a quarterly basis.
• Independent audits (internal or third-party) of the moderation pipeline, detection-tooling accuracy, and 2257 record completeness are conducted at least annually.
• Key metrics, including time-to-removal, false-negative rate on detection tooling, NCMEC report volume/turnaround, and appeal outcomes, are tracked and reviewed monthly by Trust & Safety leadership.

18. Policy Review and Amendment

This Policy is reviewed at least annually and additionally whenever there is a material change in applicable law, detection technology, or emerging harm patterns (e.g., new AI-generation methods). All amendments are version-controlled, dated, and approved by the designated Policy Owner and Legal Counsel before taking effect.