**Facts & Prevalence — Sweden (2026 Order-of-Magnitude Estimates)**

All figures below are **synthetic and illustrative only**. They follow the methodology in *Pedalgo_Prevalence_Sweden_EN.pdf* and *Pedalgo_Sweden_Facebook_Estimate.pdf*. No real platform data with reviewer-confirmed labels yet exists. These numbers serve solely to illustrate scale and triage priority for human reviewers. They are not evidence of guilt or prevalence of offences.

### Universal Base Rate Principle
A single base rate for pedophilic attraction is applied across the entire male population. This choice reflects the absence of reliable evidence for ethnicity- or group-specific weighting in clinical and epidemiological literature. Using one rate avoids introducing unsupported assumptions into risk-signal calculations.

### Key Distinctions in Scale
- **Has pedophilic attraction**: largest group.
- **Has ever committed a contact offence**: substantially smaller.
- **Currently active in detectable predatory contact on one specific platform**: orders of magnitude smaller still.

These layers are not interchangeable. Most individuals with attraction never progress to offending; most who have offended are not currently active on any given platform.

### 2026 Order-of-Magnitude Estimates (Sweden)

**Adult men with pedophilic attraction**  
- Approximate range: 25 000 – 60 000  
- Confidence: low–moderate (based on international clinical estimates applied to Swedish population projections).  
- Source principle: unchanged universal base rate; no platform or behavioural data required.

**Men who have ever committed a contact sexual offence against a child**  
- Approximate range: 3 000 – 8 000 (cumulative, not annual).  
- Confidence: low (official conviction and victim-report data are incomplete and lag).  
- This figure remains far below the attraction estimate, illustrating the conversion gap.

**Currently active in detectable predatory contact on major Swedish platforms (combined)**  
- Approximate range: low hundreds (order of magnitude).  
- Platforms considered: Facebook/Instagram, Snapchat, Discord, Roblox, TikTok (Swedish user bases).  
- Confidence: very low (no verified 2025–2026 labelled data available).  
- Only a small subset of the “ever offended” group is expected to be active on any single platform at one time.

**Platform-specific illustration (synthetic example)**  
- On a platform with ~2 million Swedish adult male users, the estimated share currently sending detectable grooming-style messages is in the low hundreds at most.  
- This is a triage signal only; human reviewers make all decisions.

### Conversion Pathway (Conceptual)
Attraction → (small fraction) ever-offending → (much smaller fraction) currently active on one platform.  
Each step multiplies the reduction in numbers. Platform activity estimates therefore remain the narrowest band and the only layer relevant to real-time detection triage.

### Confidence Ranges and Limitations
- All ranges carry wide uncertainty because they rest on pre-2025 international studies extrapolated to 2026 Swedish demographics.
- No ethnicity weighting is applied at any stage.
- Estimates will be revised only when reviewer-confirmed platform data become available.

These figures are intended exclusively for internal prioritisation of human review queues and for transparent public explanation of scale. They do not constitute accusations, risk scores for individuals, or operational thresholds. All escalation follows lawful channels (NCMEC CyberTipline, Swedish Police, IWF, DSA trusted flagger routes).