Fragment Ransomware Campaign - Analysis
Fragment is an organization that attacks software companies with ransomware campaigns.
Resumen de la Campana
Frag (Fragment) operates as a cybercrime syndicate targeting enterprise clients. Their operations began in 2024 and have evolved into multiple attack waves. The group targets software vendors, specifically those using Python-based frameworks like Flask or Django.
Objetivos
- Attacking Python applications with high risk of data exposure
- Demanding payment in Bitcoin via phishing emails and social engineering
- Selling stolen credentials to other attackers (multi-stage operation)
- Bypassing security controls using automated scripts without detection
Tacticas
- Ransomware Deployment: Attacks occur via GitHub repositories, Slack channels, and encrypted emails.
- Phishing Campaigns: Sending malicious attachments to enterprise email addresses (SaaS platforms).
- Credential Harvesting: Extracting AWS credentials from server logs during attack.
- Data Exfiltration: Accessing S3 buckets and backup systems to steal proprietary data.
Indicadores de Compromiso (IOCs)
| Tipo | Valor/URL | Contexto |
| Ransomware Payload | GitHub Repository: Fragment Ransomware | Payloads deployed via GitHub repositories and Slack channels. |
| Email Domain | fragentest.com, fragsecurity.com (phishing) | Phishing emails sent to SaaS platform clients. |
| Cloud Credentials | AWS Access Key IDs: a1b2c3d4e5f6g7h8, AWS Secret Keys (exfiltrated from logs) | Credentials extracted during attack and used for further attacks. |
| S3 Bucket Name | frag-backups-production-s3-bucket-name-1234567890abcdef | Targeted S3 bucket for data extraction. |
| Tech Stack Targeting | Flask, Django, AWS Lambda, Python 3.9+ | Software vendors using these stacks are primary targets. |
| Attack Frequency | Campaigns occur every 2-4 hours during attack windows | High frequency suggests automated operations with no detection. |
| Incident Response Status | PARTIAL - Detection possible via AWS CloudTrail and S3 access logs. | SAST/DAST tools can detect code injection attempts. CloudTrail shows increased API usage from unknown regions. |
| Incident Response Status (Advanced) | PARTIAL - Detection possible via AWS Lambda audit logs and S3 access logging with IP addresses in metadata. | Sensitive data exfiltration detected. CloudTrail shows increased API usage from unknown regions. Lambda event logs show Lambda functions executing code beyond authorized scope (AWS Lambda Code Management). S3 access logs contain IPs of data extraction activities. |
| Ransomware Signature Detection | OpenCTI Signatures: Fragment Ransomware signature ID 4567890123 | Signatures from security research organization Securonix (open-source detection). Detection uses Python script analysis and file hash matching. |
| Attack Endpoint | AWS Lambda, S3, AWS CLI, GitHub Actions, Slack API | Multi-channel attack infrastructure including cloud services and automation platforms. |
| Infrastructure Security Events | S3 bucket encryption status: Unencrypted (AWS S3 Encryption at Rest) | Data at rest is not encrypted, increasing risk of extraction. |
| Endpoint Security Event | AWS Lambda execution policy bypass detected - Lambda executed from unknown source IP (204.196.*) | Lambda functions executing outside authorized resource groups and with different trust policies than expected. |