README.md

---
license: cc-by-4.0
task_categories:
- tabular-classification
tags:
- cybersecurity
- fraud
- mobile-banking
- phishing
- fintech
- sub-saharan-africa
- synthetic
- scam
pretty_name: African Mobile Banking Phishing
size_categories:
- 10K<n<100K
language:
- en
configs:
- config_name: baseline
  data_files: data/phishing_baseline.csv
- config_name: ai_amplification
  data_files: data/phishing_ai_amplification.csv
- config_name: enhanced_protection
  data_files: data/phishing_enhanced_protection.csv
data_type: synthetic
---

> ⚠️ **Synthetic dataset** — Parameterized from published SSA literature, not real observations. Not suitable for empirical analysis or policy inference.

# African Mobile Banking Phishing

Synthetic dataset on mobile banking phishing attacks across 15 African economies. Tracks attack vectors (SMS/USSD), success rates, financial losses, victim demographics, and detection patterns. Designed for fraud detection research, cybersecurity analysis, and fintech risk assessment.

## Dataset Description

- **18,000** synthetic phishing campaign records
- **15 countries**: Major mobile money markets in Africa
- **3 scenarios**: baseline, ai_amplification, enhanced_protection
- **26 variables** per record

## Variables

| Variable | Type | Description |
|---|---|---|
| record_id | string | Unique campaign identifier |
| year | int | Year of record |
| country | string | Target country |
| region | string | Regional grouping |
| mobile_maturity_index | float | Mobile money maturity 0-1 |
| campaign_id | string | Unique campaign ID |
| attack_vector | string | sms_phishing, ussd_phishing, voice, social_media |
| target_platform | string | Target banking app |
| num_targets | int | Number of targets in campaign |
| num_successful_compromises | int | Successful infections |
| attack_success_rate_pct | float | Success rate percentage |
| total_financial_loss_usd | float | Total loss in USD |
| average_loss_per_victim_usd | float | Average victim loss |
| campaign_duration_days | float | Campaign duration |
| fraud_stage | string | Attack stage |
| message_theme | string | Phishing theme |
| sophistication_score | float | Attack sophistication 0-1 |
| target_sophistication_needed | float | Target sophistication needed |
| time_to_detect_days | float | Detection time |
| detection_rate_pct | float | Detection rate |
| user_awareness_score | float | User awareness level |
| law_enforcement_response | string | Law enforcement action |
| victim_age_group | string | Victim age distribution |
| victim_education | string | Victim education level |
| reporting_rate_pct | float | Victim reporting rate |
| repeat_victim_pct | float | Repeat victim percentage |
| cross_border_attack | int | Cross-border attack flag |
| scenario | string | baseline, ai_amplification, enhanced_protection |

## Scenarios

- **baseline** (6K): Pre-AI phishing, traditional methods, 2018-2021. SMS/USSD vectors, moderate success rates.
- **ai_amplification** (6K): AI-powered attacks, deepfakes, 2022-2024. Higher sophistication, increased losses.
- **enhanced_protection** (6K): Better detection, awareness, 2025-2026. Reduced success rates, faster detection.

## Generation Methodology

Parameters calibrated against:
- GASA State of Scams in Africa 2025
- SABRIC South Africa Crime Statistics 2024
- Kenya Banking Fraud Losses 2025 ($1.6B)
- Nature smishing attacks research
- ADF Magazine mobile money security analysis

Fraud rates based on:
- Kenya: 4.2% fraud rate, M-Pesa ecosystem
- Nigeria: 5.5% fraud rate
- South Africa: 3.8% fraud rate, 74% increase in 2025
- 68% of Africans report scam experience

## Use Cases

- Phishing detection model training
- Fraud pattern analysis
- Financial loss forecasting
- User awareness campaign planning
- Law enforcement resource allocation
- Cross-border threat intelligence

## Citation

```bibtex
@dataset{african_mobile_banking_phishing_2026,
  title={African Mobile Banking Phishing Dataset},
  author={ElectricSheepAfrica},
  year={2026},
  license={cc-by-4.0}
}
```

## License

CC BY 4.0 - This is synthetic data generated for research and educational purposes.