More than 575 malicious OpenClaw skills delivered malware
Original title: Over 575 malicious OpenClaw skills delivered malware to Windows and macOS
Acronis also describes an AI-agent operator, next-day exploitation of a patched Gitea flaw, and ransomware that removes recovery data.
Acronis TRU’s Cyber Pulse 2026 says more than 575 malicious OpenClaw skills posed as utilities and delivered trojans, cryptocurrency miners and information stealers to Windows and macOS. The report also describes one operator coordinating AI models and tools across campaigns, malicious Hugging Face payloads, and stolen browser sessions used to reach cloud services. In the Red Heron activity, a public exploit for a critical Gitea flaw appeared the day after a patch, and confirmed breaches included stolen repositories.
Unconfirmed: The report does not name the Gitea CVE or confirm how many of the described campaigns are independently verified.
- BOSSAI linked one operator to more than 10 AI models and tools for scanning and exploitation.
- Reported targeting included Taiwan government systems, Thai universities and Mexican telephone data.
- Of 1.8 million endpoints running Acronis RMM, 63.1% had another remote-access tool.
- Qilin, LockBit 5.0 and INC samples could remove shadows, snapshots or shut down virtual machines.
- TamperedChef apps carried valid Extended Validation certificates from U.S.-registered shell companies.
Impact: Malware delivery, stolen repositories and disrupted recovery
| Product | Affected |
|---|---|
| OpenClaw skills | — |
| Gitea | — |
- 1.Review OpenClaw skills before granting them agent permissions and executable access
- ATT&CK
- T1195.002 Compromise Software Supply ChainT1189 Drive-by CompromiseT1486 Data Encrypted for ImpactT1490 Inhibit System RecoveryT1562.001 Impair Defenses: Disable or Modify Tools
- Vendors
- Gitea
- Products
- OpenClawWindowsmacOSVMware ESXi
- Malware
- TamperedChefLockBit 5.0
- Organizations
- Acronis
- Countries
- TaiwanThailandMexicoUnited States
From the original article
3 min read at helpnetsecurity.com
Cybercriminals are using AI agents to coordinate attacks, stolen credentials to access linked services, and tools that interfere with intrusion detection and system recovery, according to Acronis TRU Cyber Pulse 2026.
By Anamarija Pogorelec · helpnetsecurity.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.