CVE-2025-2492 is an improper authentication vulnerability in the AiCloud functionality of certain ASUS router firmware releases. A crafted request can bypass authentication controls and potentially execute router functions without authorization. Affected firmware belongs to the 3.0.0.4_382, 3.0.0.4_386, 3.0.0.4_388, and 3.0.0.6_102 series. Routers with AiCloud enabled are affected; remotely reachable AiCloud services expose the vulnerability to network-based attacks. Potential consequences include unauthorized configuration changes and arbitrary code execution.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
1 valid exploit after Mallory filtered fakes, detection scripts, and README-only repos.
Repository contains a single substantive Go program, origasus.go, plus a README. The code is an operational exploit/scanner targeting ASUS AiCloud/AsusWRT devices and explicitly references a chained attack involving SETROOTCERTIFICATE to write /etc/cert.pem.1 and APPLYAPP/RC_SERVICE to execute commands. It is not merely a detector: it reads targets from stdin or a file, supports optional TLS and multi-port scanning, verifies ASUS-related indicators to reduce false positives, and then attempts exploitation using multiple HTTP request and shell-execution variants. The exploit is structured as a concurrent scanner with global configuration, signal handling, exploited-host tracking, and environment-driven loader customization. It maintains a list of common ASUS management ports, supports host:port parsing or separator-based input, and skips previously exploited hosts unless disabled. The payload logic is the most notable part: it builds several shell-script variants intended to be written to /etc/cert.pem.1, then tries many command-injection forms to execute that file. The staged script attempts to download kla.sh from a configurable loader host over HTTP or raw TCP using wget, busybox wget, curl, nc, or toybox nc, stores it in writable temp locations, marks it executable, and launches it in the background with a campaign tag. Fingerprintable observables include the default loader IP 11.11.11.11, HTTP path /bins/kla.sh, TCP port 3342, target-side file /etc/cert.pem.1, temp directories /dev/shm, /var/tmp, /tmp, and the local bookkeeping file exploited.txt. Overall, this repository is a compact standalone Go-based exploitation utility for mass-targeting vulnerable ASUS router/web-management interfaces, with built-in staging for a second-phase shell payload.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
36 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
A specific vulnerability in ASUS AiCloud routers that the attacker infrastructure was used to exploit in early 2026.
A vulnerability affecting ASUS AiCloud Routers that the campaign singled out as part of apparent efforts to broaden the ORB network.
A vulnerability affecting ASUS AiCloud Routers that was exploited from infrastructure linked to UAT-7810 or an associated actor to likely expand the ORB network.
A known n-day vulnerability in ASUS AiCloud routers that UAT-7810 reportedly exploits for initial access as part of expanding its ORB infrastructure.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.