Loki Intelligence — Security Briefs · Published
Daily Security Brief: MikroTik Takeovers, BIG-IP Zero-Day, and Container Escapes
Pre-patch exploitation of MikroTik routers and an F5 BIG-IP APM zero-day show attackers moving fast on edge systems, while an unpatched Ubuntu kernel flaw weakens container isolation. Teams should apply hotfixes and firmware updates, check devices for persistence, and isolate untrusted workloads until kernel patches arrive.
Signal 01 · The Hacker News
Critical Bifrost AI Gateway Flaw Lets Attackers Run Commands Without Credentials
CVE-2026-90898 in the Bifrost AI gateway lets a single unauthenticated request to the management API start arbitrary commands as the gateway user when management authentication is off, which is the default. The official Docker image binds to all interfaces, raising exposure if the port is published. Transports v2.1.0 blocks the attack.
Why it matters: AI gateways concentrate LLM provider keys; enable authentication, upgrade, avoid publishing the management port, and treat exposed unauthenticated deployments as potentially compromised.
Source: The Hacker News
Signal 02 · The Hacker News
MikroTrick Chain Let Attackers Take Over MikroTik Routers Without a Password or SSH Key
CERT Polska detailed MikroTrick, a chain of two RouterOS SSH flaws (CVE-2026-67279 and CVE-2026-86060) that grants full administrative control without valid credentials when SSH is reachable. Exploitation traces appeared a day before patches in RouterOS 6.49.21, 7.23.4 and 7.24.2, and attackers created rogue admin accounts.
Why it matters: Patch RouterOS, keep SSH off untrusted networks, and hunt for unknown accounts, scripts and scheduler entries, because patching does not undo earlier attacker changes.
Source: The Hacker News
Signal 03 · CISA
CISA Adds Four Known Exploited Vulnerabilities to Catalog
CISA's September 22 KEV update lists four exploited flaws: Check Point improper certificate validation (CVE-2026-85102) and path traversal (CVE-2026-93616), Arista VeloCloud Orchestrator improper input validation (CVE-2026-93952), and an F5 BIG-IP APM heap-based buffer overflow (CVE-2026-94127). CISA urges all organizations to prioritize them.
Why it matters: Confirm whether any of these products are internet-facing, apply vendor patches or mitigations, and preserve logs for compromise checks on security gateways.
Source: CISA
Signal 04 · The Hacker News
545 Hackers Tested It First. Now XRanges for AI Scores Your Security Agent
A partner-contributed article describes XRanges for AI from CTF.ae, a platform that deploys instrumented target applications and scores autonomous security agents on coverage, rule-of-engagement violations, confirmed exploitation and environment integrity. It says the same instrumentation monitored 545 players during a DEF CON 34 bug bounty CTF.
Why it matters: Teams adopting AI security agents should measure what the agents actually did, including rule violations and blind spots, instead of trusting self-reported findings.
Source: The Hacker News
Signal 05 · The Hacker News
Exploit Released for Unpatched Ubuntu Linux Flaw Enabling Host-Root Container Escape
DepthFirst released exploit code for CVE-2026-80521 (CVSS 7.8), a use-after-free in the Linux kernel's AF_UNIX socket garbage collector that lets code inside a container gain root on the host. The fix landed upstream on August 6, but Ubuntu 26.04, 24.04 and 22.04 kernels remain unpatched with no published date.
Why it matters: Until distribution kernels ship fixes, move untrusted workloads to microVM isolation or dedicated hosts, and watch Ubuntu's security tracker for patched packages.
Source: The Hacker News
Signal 06 · The Hacker News
F5 Patches Critical BIG-IP APM Zero-Day Exploited for Unauthenticated RCE on OAuth Servers
F5 released engineering hotfixes for CVE-2026-94127 (CVSS 9.8), an exploited heap overflow in BIG-IP APM that allows unauthenticated code execution when APM acts as an OAuth authorization server. Restricting management interface access does not help because the malicious traffic targets the virtual server. CISA added it to KEV on September 22.
Why it matters: Identify APM OAuth authorization server setups, preserve forensic evidence, apply the hotfix or F5's iRule mitigation, and review APM and audit logs for compromise indicators.
Source: The Hacker News
Signal 07 · The Hacker News
Researcher Drops BigDiskBuster Zero-Day PoC That Blocks Microsoft Defender Updates
Former Microsoft researcher Abdelhamid Naceri released BigDiskBuster, a proof-of-concept that blocks Microsoft Defender updates by briefly filling all free disk space whenever an update starts. No patch, CVE or advisory exists. His earlier Defender tools were used in real intrusions before Microsoft fixed them and CISA listed them as exploited.
Why it matters: Track Defender engine and signature versions across the fleet, alert on update failures and low disk space, and restrict execution of unknown tools.
Source: The Hacker News
Brief sources
Related briefs
- Hardware Hacking for Embedded Security
- How to Evaluate an AI Security Testing Provider
- What Is AI Agent Security Testing?
Relevant Loki services: Orvyn — AI agent security private preview and SKYEN web & API pentesting.