Skip to content
  • Home
  • Cyber Map
  • About Us – Contact
  • Disclaimer
  • Terms and Rules
  • Privacy Policy
Cyber Web Spider Blog – News

Cyber Web Spider Blog – News

Globe Threat Map provides a real-time, interactive 3D visualization of global cyber threats. Monitor DDoS attacks, malware, and hacking attempts with geo-located arcs on a rotating globe. Stay informed with live logs and archive stats.

  • Home
  • Cyber Map
  • Cyber Security News
  • Security Week News
  • The Hacker News
  • How To?
  • Toggle search form
Infostealer Targets OpenClaw AI, Exposes Security Flaws

Infostealer Targets OpenClaw AI, Exposes Security Flaws

Posted on February 16, 2026 By CWS

In a significant development in cybersecurity, experts have identified a malicious infostealer that successfully extracted sensitive configuration files from OpenClaw, an AI platform previously known as Clawdbot and Moltbot. The incident highlights a worrying trend in cyber threats, shifting from conventional credential theft to targeting the core of personal AI systems.

Infostealer’s Advanced Tactics

The attack, attributed to a variant of the well-known Vidar malware, was uncovered by Hudson Rock’s cybersecurity team. Unlike previous instances, this breach did not rely on a custom module for OpenClaw. Instead, it used a broad-spectrum file-grabbing routine to locate files with specific extensions and directory names, capturing critical data in the process.

The compromised files included openclaw.json, which holds the gateway token along with user information, device.json, containing cryptographic keys, and soul.md, detailing the AI’s operational and ethical guidelines. Such data could enable attackers to remotely access or impersonate the victim’s AI agent if network vulnerabilities are present.

Implications for AI and Cybersecurity

This breach underscores the evolving threat landscape as AI systems become more integral to professional environments. According to Hudson Rock, malware developers may soon focus on creating specialized tools to decrypt and analyze AI-related files, similar to existing tools for platforms like Chrome and Telegram.

In response to these security challenges, OpenClaw’s maintainers have partnered with VirusTotal to enhance detection and prevention measures. This collaboration aims to identify malicious skills and rectify potential misconfigurations, bolstering the platform’s defenses against similar attacks.

Broader Security Concerns

The incident also sheds light on other vulnerabilities within the ecosystem. A recent campaign detailed by OpenSourceMalware revealed that some attackers are bypassing VirusTotal scans by hosting malware externally while using decoy skills on ClawHub, further complicating detection efforts.

Additionally, research by OX Security has highlighted issues with Moltbook, a forum for AI agents, where accounts cannot be deleted once created, posing risks for data privacy. SecurityScorecard’s STRIKE team reported numerous exposed OpenClaw instances, potentially vulnerable to remote code execution (RCE) attacks, which allow unauthorized code execution on compromised systems.

Amidst these challenges, OpenClaw continues to gain traction, with over 200,000 stars on GitHub since its launch in November 2025. As OpenAI CEO Sam Altman announced the integration of its founder, Peter Steinberger, into OpenAI, the platform’s future is poised for further development, though security remains a critical concern.

The Hacker News Tags:AI agents, AI security, Cybersecurity, InfoStealer, OpenClaw, OpenClaw breach, OpenClaw tokens, RCE vulnerabilities, threat intelligence, Vidar

Post navigation

Previous Post: AI Assistants Vulnerable to Hidden Memory Manipulations
Next Post: Noodlophile Malware Uses Fake Jobs to Evade Security

Related Posts

China-Aligned Threat Group Uses Windows Group Policy to Deploy Espionage Malware China-Aligned Threat Group Uses Windows Group Policy to Deploy Espionage Malware The Hacker News
AI Is Already the #1 Data Exfiltration Channel in the Enterprise AI Is Already the #1 Data Exfiltration Channel in the Enterprise The Hacker News
New Oracle E-Business Suite Bug Could Let Hackers Access Data Without Login New Oracle E-Business Suite Bug Could Let Hackers Access Data Without Login The Hacker News
Researchers Reveal Reprompt Attack Allowing Single-Click Data Exfiltration From Microsoft Copilot Researchers Reveal Reprompt Attack Allowing Single-Click Data Exfiltration From Microsoft Copilot The Hacker News
Policy, Isolation, and Data Controls That Actually Work Policy, Isolation, and Data Controls That Actually Work The Hacker News
A Pragmatic Approach To NHI Inventories  A Pragmatic Approach To NHI Inventories  The Hacker News

Categories

  • Cyber Security News
  • How To?
  • Security Week News
  • The Hacker News

Recent Posts

  • LockBit 5.0 Targets Multiple Systems with Enhanced Ransomware
  • Cloud Password Managers Face Security Challenges
  • Noodlophile Malware Uses Fake Jobs to Evade Security
  • Infostealer Targets OpenClaw AI, Exposes Security Flaws
  • AI Assistants Vulnerable to Hidden Memory Manipulations

Pages

  • About Us – Contact
  • Disclaimer
  • Privacy Policy
  • Terms and Rules

Archives

  • February 2026
  • January 2026
  • December 2025
  • November 2025
  • October 2025
  • September 2025
  • August 2025
  • July 2025
  • June 2025
  • May 2025

Recent Posts

  • LockBit 5.0 Targets Multiple Systems with Enhanced Ransomware
  • Cloud Password Managers Face Security Challenges
  • Noodlophile Malware Uses Fake Jobs to Evade Security
  • Infostealer Targets OpenClaw AI, Exposes Security Flaws
  • AI Assistants Vulnerable to Hidden Memory Manipulations

Pages

  • About Us – Contact
  • Disclaimer
  • Privacy Policy
  • Terms and Rules

Categories

  • Cyber Security News
  • How To?
  • Security Week News
  • The Hacker News