Loki Intelligence — Security Briefs · Published
Daily Security Brief: Exploited DevOps, VoIP, AI Agent, and KEV Risks
Today’s brief centers on actively exploited enterprise software flaws, especially in artifact management and VoIP platforms, alongside CISA’s expanded KEV list. Developer environments also remain under pressure from credential-stealing malware and unsafe AI coding agent behaviors.
Signal 01 · BleepingComputer
Hackers exploit critical JFrog Artifactory flaw to forge admin tokens
Attackers are abusing a critical JFrog Artifactory authentication bypass to create administrative access tokens. Organizations using Artifactory should treat exposed or unpatched instances as high-priority risk.
Why it matters: Patch quickly, review token creation activity, rotate sensitive credentials, and validate that administrative access is tightly monitored.
Source: BleepingComputer
Signal 02 · BleepingComputer
Hackers exploit Sangoma Switchvox flaw to deploy reverse shells
Sangoma Switchvox systems are being targeted through a critical unauthenticated SQL injection flaw that can lead to remote code execution. The issue affects VoIP infrastructure that may be internet-facing and operationally sensitive.
Why it matters: Prioritize patching, restrict external access to VoIP admin surfaces, and hunt for abnormal process execution or outbound connections from PBX servers.
Source: BleepingComputer
Signal 03 · The Hacker News
Researchers Use Claude to Port Pre-Auth RCE Exploit From One PLC Model to Another
Researchers reported using an AI assistant to adapt prior PLC vulnerability research from one industrial controller model to another. The work highlights how AI can accelerate vulnerability analysis across similar embedded systems.
Why it matters: Asset owners should not assume variant devices are safe once a related model is vulnerable; segment OT networks, apply vendor fixes, and monitor controller behavior.
Source: The Hacker News
Signal 04 · The Hacker News
Attackers Exploit Critical Switchvox Flaw to Deploy Reverse Shells Without Credentials
The Hacker News published a relevant security update.
Why it matters: Review exposure, ownership, and remediation priority for systems that match this signal.
Source: The Hacker News
Signal 05 · CISA
CISA Adds Seven Known Exploited Vulnerabilities to Catalog
CISA added seven vulnerabilities to its Known Exploited Vulnerabilities catalog, including flaws affecting Switchvox, Starlette, Kestra OSS, and other products. The additions confirm active abuse and create clear remediation deadlines for many organizations.
Why it matters: Use the KEV update to drive patch prioritization, asset owner outreach, compensating controls, and evidence-based executive risk reporting.
Source: CISA
Signal 06 · The Hacker News
Shai-Hulud's Reach Just Grew to 469 Credential Locations. Here's What That Means
A newer Shai-Hulud infostealer variant reportedly searches hundreds of developer, CI/CD, cloud, and AI tool locations for secrets. Its broadened targeting reflects how valuable engineering workstations and automation environments have become.
Why it matters: Reduce local secret storage, enforce short-lived credentials, monitor for unusual repository or CI activity, and rotate exposed tokens rapidly.
Source: The Hacker News
Signal 07 · The Hacker News
Malicious .git Configs Can Make Claude, Codex, Cursor, and Other AI Agents Run Attacker Code
Security researchers disclosed flaws in multiple command-line AI coding agents where repository-controlled Git settings could trigger unintended command execution. Several issues were reportedly unresolved at publication.
Why it matters: Treat unfamiliar repositories as untrusted, isolate AI coding tools, disable risky automation defaults, and require agent vendors to document secure execution boundaries.
Source: The Hacker News
Brief sources
Related briefs
- What Is AI Agent Security Testing?
- AI Red Teaming vs Traditional Pentesting
- MCP Security Testing Checklist
Relevant Loki service: Orvyn — AI agent security private preview.