AI Content Moderation

AI → Human

AI auto-reviews user content; flagged items go to human moderators.

5 nodes · 4 edgessocial

agenthumansystem

Visual

Content Submittedevent

User posts text, image, or video.

↓sequential→ AI Safety Scan

AI Safety Scanagent

Check for hate speech, NSFW, spam, misinformation.

↓conditional→ Auto-Approve

↓conditional→ Human Moderator Review

Auto-Approvesystem

Content passes all checks.

Human Moderator Reviewhuman

Moderator reviews flagged content.

↓sequential→ Enforce Decision

Enforce Decisionsystem

Publish, remove, or restrict content.

Open in Editor

uc-content-moderation.osop.yaml

osop_version: "1.0"
id: "content-moderation"
name: "AI Content Moderation"
description: "AI auto-reviews user content; flagged items go to human moderators."

nodes:
  - id: "content_submitted"
    type: "event"
    name: "Content Submitted"
    description: "User posts text, image, or video."

  - id: "ai_scan"
    type: "agent"
    subtype: "llm"
    name: "AI Safety Scan"
    description: "Check for hate speech, NSFW, spam, misinformation."
    security:
      risk_level: "low"

  - id: "auto_approve"
    type: "system"
    name: "Auto-Approve"
    description: "Content passes all checks."

  - id: "flag_review"
    type: "human"
    subtype: "review"
    name: "Human Moderator Review"
    description: "Moderator reviews flagged content."
    security:
      approval_gate: true

  - id: "action"
    type: "system"
    name: "Enforce Decision"
    description: "Publish, remove, or restrict content."

edges:
  - from: "content_submitted"
    to: "ai_scan"
    mode: "sequential"
  - from: "ai_scan"
    to: "auto_approve"
    mode: "conditional"
    when: "scan.safe == true"
  - from: "ai_scan"
    to: "flag_review"
    mode: "conditional"
    when: "scan.safe == false"
  - from: "flag_review"
    to: "action"
    mode: "sequential"

←

AI Anomaly Detection & Human Triage

compliance-audit

→