**SYNTHETIC — FOR ILLUSTRATION ONLY — NOT REAL CASES**

All scenarios below are entirely invented for educational purposes. They demonstrate how a Pedalgo-style triage system could surface conversations for human review. Scores are risk signals only; they never determine guilt and always result in human evaluation. Lawful escalation occurs exclusively through official channels (NCMEC CyberTipline, Swedish Police, IWF, or DSA trusted flagger routes).

### Example 1 — Low Risk (Proactive)
**Excerpt**  
Adult: “Hi, I saw your post about the new game. How old are you?”  
Minor: “13”  
Adult: “Cool, my niece is also into that game.”

**Signals detected**  
Single age inquiry followed by an immediate, non-sexual reference to family. No secrecy requests, no repeated contact attempts, no sexual framing.

**Composite score band**  
Low (illustrative range 0–25).

**Reviewer action**  
Queue for routine spot-check. No escalation required unless additional context appears in follow-up messages.

### Example 2 — Low Risk (Reactive)
**Excerpt**  
Minor posts about school stress. Adult replies: “That sounds tough. Talk to a teacher or your parents—they can help.”

**Signals detected**  
Supportive language with no age-focused probing or private-channel suggestions.

**Composite score band**  
Low (0–25).

**Reviewer action**  
No action beyond standard logging. Content does not meet review threshold.

### Example 3 — Medium Risk (Proactive)
**Excerpt**  
Adult: “You seem mature for your age. Don’t tell your parents we’re chatting—it’s our secret.”  
Minor: “Ok.”

**Signals detected**  
Age-related compliment combined with explicit request for secrecy and concealment from guardians.

**Composite score band**  
Medium (illustrative range 26–65).

**Reviewer action**  
Prioritize for human review within 24 hours. Reviewer assesses full thread and context before any decision on reporting.

### Example 4 — Medium Risk (Reactive)
**Excerpt**  
Minor shares feeling lonely. Adult: “I can be your special friend. Send me a picture of yourself so I know what you look like.”

**Signals detected**  
Rapid shift to “special friend” framing plus request for personal imagery without clear malicious intent visible in this fragment.

**Composite score band**  
Medium (26–65).

**Reviewer action**  
Human reviewer examines conversation history and platform metadata. Decision on forwarding to authorities remains with the reviewer.

### Example 5 — High Risk (Proactive)
**Excerpt**  
Adult: “I can send you money if you keep this between us and meet me after school. Just don’t say anything to anyone.”

**Signals detected**  
Offer of material benefit, insistence on secrecy, and suggestion of offline meeting with a minor.

**Composite score band**  
High (illustrative range 66–100).

**Reviewer action**  
Expedited human review (within hours). If confirmed as high-risk, route directly to Swedish authorities or NCMEC CyberTipline per established protocol. No automated action taken.

### Example 6 — High Risk (Reactive)
**Excerpt**  
Minor describes discomfort with previous messages. Adult replies: “Ignore them. Send me the photos we talked about and I’ll delete everything after.”

**Signals detected**  
Coercive language, reference to prior image requests, and instruction to conceal activity.

**Composite score band**  
High (66–100).

**Reviewer action**  
Immediate human review. Lawful report prepared for appropriate authorities; platform safety team notified through official channels only.

### Example 7 — Medium Risk (Proactive, ambiguous grooming attempt)
**Excerpt**  
Adult: “You look really pretty in your profile picture. Want to chat privately where no one else can see?”

**Signals detected**  
Appearance-focused comment plus move to private channel with emphasis on exclusivity.

**Composite score band**  
Medium (26–65).

**Reviewer action**  
Standard human review queue. Context from surrounding messages determines whether score is adjusted.

These examples illustrate triage prioritization only. All final decisions rest with trained human reviewers following Swedish and EU legal frameworks.