← Back to Feed CACHED · 2026-09-04 17:35:42 · CACHE_KEY CVE-2026-75754
CVE-2026-75754 · CWE-306 · Disclosed 2026-09-04

Missing Authentication for Critical Function

ASSESSED — NOISGATE V0.5
Vendor
Reassessed
Verdict:
Do you agree?
01 · The Real Story

It's like leaving the master key to every office in the building taped to the lobby wall, with the alarm code written on the back

CVE-2026-75754 chains three flaws in ASUS Control Center Enterprise (ACC) — the centralized management platform that controls firmware updates, monitoring, and remote operations across your entire ASUS server and workstation fleet. An unauthenticated attacker sends a single HTTP request to extract the ACC encryption key (CWE-306), leverages SSRF (CWE-918) to instruct a local service to enable SSH on port 2222, then logs in with hardcoded credentials (CWE-798) for a root shell. From there, the attacker has direct read/write/delete access to ACC data and remote control over every server, PC, and workstation the platform manages. All ACC versions appear affected; no patched version has been publicly confirmed as of 2026-09-05.

There is no official CVSS 3.1 score from ASUS or NVD yet, though a CVSS 4.0 vector of 10.0 has been published on threat-intelligence aggregators. A perfect 10 is warranted on paper — network-reachable, zero-auth, zero-complexity, full CIA impact with downstream scope — but it overestimates real-world exposure slightly because ACC management interfaces are typically internal-only. That said, this is a fleet-orchestration platform: compromise here IS fleet compromise. The severity is accurate for any environment where ACC is reachable from any untrusted network segment, and still extremely dangerous even when restricted to the management VLAN because any lateral-movement foothold reaches it.

"Unauthenticated chain to root on your fleet management console — game over for every managed host"
02 · The Attack Path

5 steps from start to impact.

STEP 01

Unauthenticated encryption key retrieval

The attacker sends a crafted HTTP request to the ACC web interface, which responds with the platform's encryption key. No authentication, no session token, no API key is required. The endpoint is accessible to anyone who can reach the ACC management port.
Conditions required:
  • Network access to ACC management interface (default HTTP/HTTPS port)
Where this breaks in practice:
  • ACC management interface is typically bound to internal/management network, not internet-facing
  • Firewall rules or VLAN segmentation may restrict access to the management port
Detection/coverage: Network IDS signatures for anomalous HTTP requests to ACC key-retrieval endpoints; web server access logs showing unauthenticated requests to sensitive API paths
STEP 02

SSRF triggers SSH enablement

Using the retrieved encryption key, the attacker crafts a server-side request that instructs an internal ACC service to enable the SSH daemon on port 2222. This is a classic SSRF — the attacker makes the server call itself to toggle a service that should never be externally controllable. The SSH service starts listening immediately.
Conditions required:
  • Valid encryption key obtained from step 1
  • ACC internal service endpoint is reachable via SSRF
Where this breaks in practice:
  • Host-based firewalls might block new port 2222 listeners, but ACC likely has permissive rules for its own services
Detection/coverage: Port scan detection for new listener on 2222; process monitoring for unexpected sshd spawn; firewall logs showing newly opened port
STEP 03

Login with hardcoded credentials

The attacker connects to the newly enabled SSH service on port 2222 and authenticates using credentials hardcoded in the ACC application. These credentials are static across all ACC installations — every deployment shares the same username and password. Authentication succeeds deterministically.
Conditions required:
  • SSH service running on port 2222 from step 2
  • Knowledge of hardcoded credentials (embedded in ACC binaries, extractable via reverse engineering)
Where this breaks in practice:
  • Credentials are not yet publicly disclosed, but are deterministic and identical across all installs — once one researcher publishes them, every ACC instance is vulnerable
Detection/coverage: SSH login monitoring on non-standard port 2222; failed/successful auth alerts from ACC host
STEP 04

Root shell on ACC server

The hardcoded credentials grant root-level access to the ACC operating environment. The attacker now has full filesystem access, can modify ACC configuration, exfiltrate managed device inventories, and execute arbitrary commands on the ACC server itself.
Conditions required:
  • Successful SSH authentication from step 3
Where this breaks in practice:
  • None — this is the natural consequence of step 3 succeeding
Detection/coverage: EDR/HIDS on ACC server detecting interactive root shell; anomalous process trees under sshd
STEP 05

Fleet-wide compromise via ACC management functions

With root access to ACC, the attacker leverages ACC's built-in fleet management capabilities — firmware push, remote command execution, task scheduling, software deployment — to pivot to every managed server, PC, and workstation. ACC is designed to manage up to 1,000+ systems per instance, so the blast radius is the entire managed fleet. The attacker can push malicious firmware, deploy ransomware, exfiltrate data, or establish persistent backdoors across all managed endpoints.
Conditions required:
  • Root access on ACC server from step 4
  • ACC managing active fleet of devices
Where this breaks in practice:
  • Managed devices must be online and reachable from ACC
  • Some organizations may have additional approval workflows for firmware pushes, but ACC root bypasses these
Detection/coverage: Anomalous bulk management operations in ACC audit logs; unexpected firmware or software deployment events; managed endpoints reporting unexpected changes
03 · Intelligence Metadata

The supporting signals.

In-the-wild exploitationNo confirmed in-the-wild exploitation as of 2026-09-05. Not listed on CISA KEV. Disclosed yesterday — exploitation window is freshly open.
Proof of conceptNo public PoC or exploit code identified yet. The hardcoded credentials have not been publicly disclosed. However, the attack chain is simple enough that weaponization is trivial once credentials are extracted from any ACC installation.
EPSS score0.00215 (approximately 78th–82nd percentile). Low probability currently, but this reflects the newness of the CVE — expect rapid climb if a PoC drops.
KEV statusNot listed as of 2026-09-05. Given the severity and trivial chain, KEV listing is plausible if exploitation is observed.
CVSS vectorCVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H — 10.0 CRITICAL (CVSS 4.0, per threat intel aggregators). No official CVSS 3.1 vector from NVD/ASUS yet.
Affected versionsAll versions of ASUS Control Center Enterprise (ACC) appear affected. No specific version range boundaries documented.
Fixed versionNo patched version confirmed as of 2026-09-05. ASUS security advisory page does not yet list a fix. Check ASUS Security Advisory for updates.
Scanning / exposure dataNo Shodan/Censys/GreyNoise data specific to ACC exposure. ACC management interfaces are typically internal-only, but deployments in flat networks or with VPN access are reachable post-initial-access.
Disclosure date2026-09-04 (published and last updated)
Researcher / orgNiels Teusink of Eye Security (Netherlands-based MSSP)
04 · The Call

noisgate verdict.

Final Verdict
= UNCHANGED to CRITICAL (9.5/10)

The single most decisive factor is role multiplier — ACC is by definition a fleet-orchestration platform, and root compromise equals remote control over every managed server and workstation. The unauthenticated, zero-complexity chain with deterministic hardcoded credentials eliminates all attacker friction, making the blast radius the entire managed infrastructure.

HIGH Vulnerability existence and attack chain mechanics
MEDIUM Exact affected version boundaries and patch availability
LOW Internet-facing exposure population (likely very small, but internal exposure is the real threat)

Why this verdict

  • Zero-auth, zero-complexity chain: The entire path from network access to root shell requires no credentials, no user interaction, and no special conditions — just HTTP access to the ACC management port. Each step follows deterministically from the previous one.
  • Hardcoded credentials eliminate probabilistic barriers: Unlike brute-force or credential-stuffing chains, the hardcoded credentials are identical across every ACC installation. Once published, exploitation is deterministic and automated.
  • Role multiplier: ACC is canonically a fleet-orchestration and infrastructure-management platform. 100% of ACC installations occupy the high-value 'orchestration/management' role by definition. Root on ACC = remote code execution on every managed endpoint. The blast radius is fleet-scale (up to 1,000+ managed devices per instance). This sets the verdict floor at CRITICAL per the deployment-role blast-radius rule.
  • No patch available: As of 2026-09-05, no fixed version exists. The vulnerability window is open-ended until ASUS ships a patch, increasing urgency.
  • Friction is minimal: The only real friction is network reachability — ACC is typically internal-facing. However, 'requires internal network access' means post-initial-access (the attacker is already inside), and management VLANs in many SMB/mid-market environments are flat or weakly segmented. This friction does not break the CRITICAL floor.

Why not higher?

This is already assessed at CRITICAL (9.5/10.0). The 0.5 reduction from a perfect 10 reflects: (1) no public PoC or hardcoded credential disclosure yet, meaning immediate mass exploitation is unlikely in the first 48 hours; (2) ACC is not typically internet-facing, so the attack requires internal network position or prior compromise; (3) ACC's installed base is smaller than mass-market products like Exchange or vCenter.

Why not lower?

Downgrading below CRITICAL is not defensible. ACC is a fleet-management platform — compromise IS fleet compromise by design. The chain requires zero authentication and uses hardcoded credentials that are identical across all installations. Even with internal-network-only exposure, any post-initial-access attacker (phishing foothold, VPN compromise, insider) reaches ACC trivially in weakly segmented environments. The blast radius is not one host — it is every managed host.

05 · Compensating Control

What to do — in priority order.

  1. Isolate ACC management interface to a dedicated, hardened management VLAN immediately — ACC should only be reachable from jump hosts / PAWs used by authorized infrastructure admins. Deploy firewall rules blocking all other subnets from reaching ACC's HTTP/HTTPS and port 2222 within the noisgate mitigation SLA of 3 days. This is the single most impactful control — if attackers can't reach the interface, the chain breaks at step 1.
  2. Block port 2222 outbound and inbound on ACC host firewall — The SSRF enables SSH on port 2222. A host-based firewall rule explicitly blocking port 2222 (both listen and connect) prevents the attacker from reaching the SSH service even if the SSRF succeeds. Deploy within 3 days.
  3. Deploy network IDS rules for ACC key-retrieval endpoint access — Monitor for HTTP requests to the ACC encryption-key endpoint. Any request to this path from a non-admin source is a high-fidelity indicator of exploitation. Alert and block.
  4. Audit ACC access logs daily until patch is available — Review ACC web server logs and SSH auth logs for anomalous access. Look for unauthenticated API calls, SSH connections on port 2222, and unexpected management operations.
  5. Evaluate ACC necessity and consider temporary shutdown — If ACC is not mission-critical for daily operations, taking it offline entirely eliminates the attack surface. This is the most aggressive but most effective control. Re-enable only after ASUS ships a patch.
What doesn't work
  • WAF in front of ACC — The key-retrieval endpoint is a legitimate HTTP path; a WAF without a custom rule for this specific endpoint won't block it. Generic OWASP rules do not cover application-specific unauthenticated API endpoints.
  • Changing SSH port — The SSRF instructs the internal service to open SSH on 2222 specifically; changing the default SSH port on the host does not affect this because the vulnerability spawns its own SSH listener.
  • Rotating credentials in ACC UI — The hardcoded credentials are embedded in the application binary, not configurable via the management interface. You cannot change them without a vendor patch.
06 · Verification

Crowdsourced verification payload.

Run this script on the ACC server itself as root or with sudo. It checks whether the SSH service is already listening on port 2222 (indicating possible exploitation) and attempts to identify whether the ACC installation is present. Invoke with: sudo bash check_cve_2026_75754.sh

noisgate-verify.sh
BASHREAD-ONLYSAFE
#!/bin/bash
# CVE-2026-75754 Checker for ASUS Control Center Enterprise
# Run on the ACC host as root: sudo bash check_cve_2026_75754.sh
# Exit codes: 0 = VULNERABLE, 1 = PATCHED, 2 = UNKNOWN

set -euo pipefail

VULNERABLE=0
PATCHED=0
UNKNOWN=0

echo "=== CVE-2026-75754 Assessment ==="
echo "Date: $(date -u +%Y-%m-%dT%H:%M:%SZ)"
echo ""

# Check if ACC is installed
ACC_FOUND=0
if [ -d "/opt/asus" ] || [ -d "/opt/ASUS" ] || [ -d "/usr/local/asus-control-center" ]; then
    ACC_FOUND=1
    echo "[INFO] ASUS Control Center installation directory detected."
elif systemctl list-units --type=service 2>/dev/null | grep -qi 'asus\|acc'; then
    ACC_FOUND=1
    echo "[INFO] ASUS Control Center service detected."
elif docker ps 2>/dev/null | grep -qi 'asus\|acc'; then
    ACC_FOUND=1
    echo "[INFO] ASUS Control Center container detected."
else
    echo "[INFO] No ASUS Control Center installation detected on this host."
    echo "RESULT: UNKNOWN (ACC not found on this host)"
    exit 2
fi

# Check if port 2222 is already listening (indicator of active exploitation)
if ss -tlnp 2>/dev/null | grep -q ':2222 '; then
    echo "[ALERT] Port 2222 is actively listening! Possible exploitation in progress."
    LISTENER_PID=$(ss -tlnp | grep ':2222 ' | grep -oP 'pid=\K[0-9]+')
    if [ -n "$LISTENER_PID" ]; then
        echo "[ALERT] Listener PID: $LISTENER_PID"
        echo "[ALERT] Process: $(ps -p $LISTENER_PID -o comm= 2>/dev/null || echo 'unknown')"
    fi
    echo ""
fi

# Since no patched version is known yet, all ACC installs are vulnerable
if [ "$ACC_FOUND" -eq 1 ]; then
    echo "[VULN] ASUS Control Center is installed and no patched version is available."
    echo "[VULN] This installation is affected by CVE-2026-75754."
    echo "[VULN] Ensure the management interface is network-isolated and port 2222 is blocked."
    echo ""
    echo "RESULT: VULNERABLE"
    exit 0
fi

echo "RESULT: UNKNOWN"
exit 2
07 · Bottom Line

If you remember one thing.

TL;DR
This is a day-one CRITICAL with no vendor patch available. Monday morning (today), your first action is network isolation: move ACC behind a dedicated management VLAN accessible only from hardened jump hosts, and deploy host firewall rules blocking port 2222. This satisfies the noisgate mitigation SLA of ≤ 3 days for CRITICAL findings. Simultaneously, open a case with ASUS support demanding an ETA for a patched ACC version — the noisgate remediation SLA is ≤ 90 days, but given the trivial unauthenticated chain and fleet-scale blast radius, you should apply the patch within days of release, not weeks. If ACC is non-essential, take it offline entirely until a fix ships. Audit ACC logs immediately for any signs of exploitation: look for HTTP requests to key-retrieval endpoints and SSH connections on port 2222. Brief your SOC that any alert from the ACC management segment is now a P1.

Sources

  1. OffSeq Threat Radar — CVE-2026-75754 Advisory
  2. THREATINT — CVE-2026-75754
  3. ASUS Security Advisory Page
  4. ASUS Control Center Data Center Edition
  5. ASUS Control Center Express — Product Overview
  6. ASUS Software-Defined Infrastructure for AI Data Centers
Peer Review

What defenders are saying.

Submit a review attribution: handle + country only
0 flags selected · stored anonymously
Validation Results

Crowdsourced verification outputs.

Results submitted by users who ran the verification payload against their environment.