xmrig

Attackers have developed an autonomous offensive agent named 'Skynet' that hijacks exposed large language model (LLM) endpoints to conduct self-directed cyberattacks. The agent leverages stolen or unsecured AI model access to perform internet-wide scanning, exploit web frameworks, steal credentials and databases, and deploy cryptocurrency miners without human intervention. It uses sophisticated jailbreak techniques to bypass commercial model safety controls by redefining malicious activities as routine development tasks. The activity represents a novel form of LLMjacking, where attackers shift computational costs to victims while maintaining anonymity and evading provider bans.

AhsayCBS, a backup solution by Singapore-based Ahsay, contains two vulnerabilities, CVE-2026-105133 and CVE-2026-105134, that have been actively exploited to achieve unauthenticated remote code execution. Threat actors have chained these flaws to deploy web shells and XMRig cryptocurrency miners on exposed systems. Initial patches in versions 9.15.2.59 and 10.3.4.0 were incomplete, and Ahsay has since released hotfixes in versions 9.15.2.70 and 10.3.4.45 to fully address the issues. The U.S. CISA has added the vulnerabilities to its Known Exploited Vulnerabilities catalog, mandating federal agencies to patch by October 17, 2026.

Threat actors are actively exploiting two unpatched vulnerabilities in the AhsayCBS backup management platform to gain unauthorized access, deploy webshells, and mine cryptocurrency. The attack chain begins with CVE-2026-105133, an authentication bypass, followed by CVE-2026-105134, which enables OS command injection. Attackers deploy JSP webshells, install the XMRig miner disguised as edge.exe, and use a PowerShell script to hide mining activity from Task Manager. The malware persists via a malicious service named 'MicrosoftEdgeUpdateSvc' and may leverage the WinRing0x64.sys driver to enhance mining performance.

Threat actors are exploiting two vulnerabilities, CVE-2026-105133 and CVE-2026-105134, in the AhsayCBS backup utility to gain remote code execution on affected systems. The attackers chain the flaws to bypass authentication and execute arbitrary commands, leading to the deployment of web shells and XMRig cryptocurrency miners disguised as Microsoft Edge. Post-exploitation activities include reconnaissance, cryptomining, and the use of an AI-assisted PowerShell script that interferes with Windows Task Manager to avoid detection.

Cybersecurity firm Huntress identified worm-like activity abusing ConnectWise ScreenConnect to propagate a four-stage VBScript chain across newly connected hosts. The attack uses social engineering, phishing, or fake refund forms to deploy rogue ScreenConnect clients, which then execute a sequence of malicious VBScripts (1.vbs to 4.vbs) for reconnaissance, payload retrieval, and execution. The final stage deploys backdoors, privilege escalation tools, or cryptocurrency miners based on system state, and the infected host re-infects new connections, creating self-propagating behavior. ConnectWise issued an advisory recommending disabling file transfer permissions to mitigate the risk.

Aeternum is a recently discovered C++ botnet loader that uses the Polygon blockchain for decentralized command-and-control (C2) operations. The malware retrieves encrypted or plaintext instructions from smart contracts on the blockchain by querying public RPC endpoints, enabling resilient and low-cost infrastructure resistant to takedowns. Three distinct malware samples were analyzed: the initial Aeternum loader, a Python-based downloader, and a multi-component payload combining XWorm RAT, XMRig cryptocurrency miner, and data exfiltration tools. The threat leverages Telegram APIs and GitHub repositories for secondary C2 and payload delivery, while using weak encryption schemes that allow decryption via contract address and payload analysis.

CISA has added seven actively exploited vulnerabilities to its Known Exploited Vulnerabilities (KEV) catalog, including critical flaws in SonicWall SMA 1000, Sangoma Switchvox, JFrog Artifactory, Kludex Starlette, Kestra OSS, and Berri LiteLLM. Threat actors are exploiting these vulnerabilities to deploy reverse shells, execute arbitrary code, steal credentials, and deploy cryptocurrency miners. Exploitation of CVE-2026-83548 and CVE-2026-83549 in SonicWall devices has been confirmed, while CVE-2026-9586 and CVE-2026-82329 are being used to gain administrative access and conduct post-exploitation activities. Microsoft and Wiz report active exploitation of CVE-2026-42271 and CVE-2026-48710 in LiteLLM deployments, with attackers achieving remote code execution and stealing API keys, and CVE-2026-49869 in Kestra being used to establish reverse shells and deploy miners.

Wiz Threat Research observed 90 days of sustained attacks against AI infrastructure through honeypots deployed across services like LiteLLM, Flowise, and LangChain. Attackers exploited vulnerabilities in MCP servers, including authentication bypass and command injection (CVE-2026-59822, CVE-2026-42271), to achieve remote code execution and deploy cryptominers. A second pattern involved blind prompt injection against AI agent frameworks, where attackers used out-of-band DNS callbacks to confirm execution and later fetch payloads from Pastebin. Post-exploitation activity was tailored to AI environments, including in-memory extraction of LiteLLM master keys, model enumeration, and use of environment-specific camouflage to hide malicious binaries.

Huntress investigated a real-world incident in June 2026 where an attacker gained initial access via a SQL injection vulnerability on a web server. After entry, the attacker conducted reconnaissance, created a backdoor user, enabled Remote Desktop, disabled Windows Defender, and deployed multiple payloads including the BadIIS malware and the XMRig cryptocurrency miner. The attacker used PowerShell scripts to maintain persistence and evade detection, highlighting the importance of not only removing malware but also identifying and patching the initial vulnerability to prevent reinfection.

Threat actors are exploiting Steam discussion forums in a social engineering campaign known as ClickFix, where they pose as helpful users offering technical fixes. They trick victims into running malicious PowerShell commands that download and execute an XMRig cryptominer. The script masquerades as a Windows optimization tool, performs fake maintenance tasks, and establishes persistence via scheduled tasks and Defender exclusions, ultimately leading to cryptocurrency mining on compromised systems.