Someone left the master key under the mat, but the mat is inside a locked maintenance closet
CVE-2026-79844 is an authentication bypass (CWE-288) in HPE Aruba Networking AOS-10, the operating system powering Aruba wireless Access Points, Gateways, and Mobility Conductors. The flaw lives in the PAPI protocol (Aruba's AP management channel on UDP port 8211). An unauthenticated attacker who can reach that port can send crafted PAPI packets to bypass authentication controls and execute arbitrary commands on the underlying host — with one caveat HPE keeps vague: *"certain preconditions outside of the attacker's control must be met."* Affected version ranges span AOS-10 10.3.0.0–10.4.1.13, 10.6.0.0–10.7.2.6, and 10.8.0.0–10.8.1.1. Patched releases are 10.4.1.14, 10.7.2.7, and 10.8.1.2 per advisory HPESBNW05162. This is one of eight CVSS 10.0 bugs in a 72-vulnerability mega-patch from HPE.
HPE scored this a perfect 10.0 — the maximum possible — which overstates the real-world risk for most enterprises. The CVSS vector (AV:N/AC:L/PR:N/UI:N/S:C) models a trivially exploitable, internet-facing, zero-click bug with scope change. But PAPI is a management-plane protocol that should never be routable from untrusted networks. In a properly segmented deployment, only the controller management VLAN can reach UDP 8211 on an AP. Censys data from prior PAPI disclosure cycles found only ~180 hosts with exposed ArubaOS management interfaces internet-wide. Combine that with HPE's own disclosure that unspecified preconditions further limit exploitability, and the 10.0 label is misleading for patch-priority triage. The bug is still dangerous — if an attacker is already on the management network, this is a one-packet-to-RCE path on every AP and gateway in range — but it is not the internet-facing zero-click apocalypse the score implies.
4 steps from start to impact.
Gain access to the PAPI management network
- Network path to target device's UDP port 8211
- Target device running a vulnerable AOS-10 version
- PAPI is a management protocol — properly segmented deployments restrict it to a management VLAN
- Internet exposure is extremely low: ~180 hosts globally per Censys historical scans
- Enterprise firewalls and ACLs typically block non-standard UDP ports from user segments
port:8211 product:ArubaOS.Send crafted PAPI packets to bypass authentication
- Network reachability to UDP 8211 (from step 1)
- Unknown environmental preconditions met on the target
- Preconditions are unspecified — attacker cannot guarantee they are met
- No public PoC or exploit tooling exists to automate this step
- PAPI is a proprietary binary protocol with limited public documentation, raising the skill bar
Execute arbitrary commands as privileged user
- Successful authentication bypass (step 2)
- Post-exploitation on embedded AOS devices requires familiarity with the proprietary filesystem and services
- EDR/AV cannot run on AOS devices — detection depends on network monitoring
- Persistence mechanisms on APs may not survive firmware upgrades
Lateral movement and infrastructure compromise
- Command execution achieved on a Conductor/Gateway (highest impact) or AP (moderate impact)
- Network segmentation between management and production limits pivot paths
- Wireless traffic may be encrypted end-to-end (WPA3-Enterprise, TLS) limiting passive interception value
- Fleet-wide push requires Conductor-level compromise, not just a single AP
cluster-security to encrypt and authenticate PAPI communications. While this does not directly mitigate CVE-2026-79844 on AOS-10, it protects AOS-8 devices from the related batch of 72 vulnerabilities and reduces the overall PAPI attack surface in mixed fleets. Deploy within 30 days.- WAF / reverse proxy — PAPI is a UDP-based binary protocol, not HTTP. Web application firewalls provide zero coverage.
- EDR / endpoint agents — AOS-10 is an embedded Linux OS on access points and gateways. No EDR agent can be installed on these devices. Detection must be network-based.
- WPA3 or 802.1X hardening — These protect wireless client authentication but do not affect the PAPI management-plane attack vector. The vulnerability is in device management, not client-facing wireless security.
- TLS/certificate pinning on PAPI — AOS-10 does not support authenticated/encrypted PAPI natively (unlike AOS-8 cluster-security). You cannot encrypt your way out of this; block at the network layer.
The supporting signals.
| In-the-Wild Exploitation | Not observed. No CISA KEV listing, no vendor confirmation of active exploitation, no campaign reporting as of 2026-10-23. HPE's advisory does not flag active exploitation. |
|---|---|
| Proof-of-Concept | None publicly available. No PoC on pocindex.io, GitHub, ExploitDB, or Nuclei templates. PAPI is a proprietary binary protocol with limited public tooling, which raises the bar for independent exploit development. |
| EPSS Score | Not yet scored. CVE was published 2026-10-22 — EPSS scoring typically lags by several days. Expect a moderate-to-high initial EPSS score given the CVSS 10.0 base and network attack vector, but the management-plane-only access pattern historically tempers EPSS predictions for PAPI bugs. |
| KEV Status | Not listed as of 2026-10-23. Prior Aruba PAPI CVEs (e.g., CVE-2024-42509, CVE-2024-47460) were also not added to KEV despite CVSS 9.8 scores, suggesting CISA considers the management-plane restriction a mitigating factor. |
| CVSS Vector Analysis | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:H/I:H/A:N — Network/Low-complexity/No-privileges/No-interaction with Scope Change. The S:C (scope change) is accurate: compromise of a Conductor changes the security scope of all managed APs. However, AV:N overstates real reachability since PAPI is a management-plane protocol. |
| Affected Versions | AOS-10: 10.3.0.0–10.4.1.13, 10.6.0.0–10.7.2.6, 10.8.0.0–10.8.1.1. Note: AOS-10 branches 10.5.x and 10.6.x-below-10.6.0.0 are end-of-maintenance and will NOT receive patches. |
| Fixed Versions | AOS-10 10.4.1.14, 10.7.2.7, 10.8.1.2. AOS-8 users should also update to 8.13.3.1 or 8.10.0.24 for the broader 72-CVE batch. |
| Scanning / Internet Exposure | Censys historically identified ~180 hosts with exposed ArubaOS management interfaces. PAPI on UDP 8211 has minimal internet exposure. Most deployments restrict PAPI to management VLANs. Shodan dork: port:8211. |
| Vendor Market Share | HPE Aruba holds approximately 11–14% of the enterprise wireless AP market (IDC 2025–2026). This means the vulnerability affects a significant installed base across Fortune 500 and mid-market enterprises. |
| Disclosure Timeline | Published 2026-10-22 per NVD/HPE. Part of a 72-vulnerability disclosure in advisory HPESBNW05162. Canadian Cyber Centre issued advisory AV26-1046 on 2026-10-15 (advance notification). |
Sources.
- HPE Advisory HPESBNW05162 — AOS-8/AOS-10 Multiple Vulnerabilities
- SecurityOnline — HPE Patches 72 Flaws in AOS-8 and AOS-10
- Strix AI — CVE-2026-79844 Detail
- VulDB — CVE-2026-79844 Entry
- Censys — ArubaOS Exposure Analysis (prior PAPI CVEs)
- HPE Support — UDP 8211 PAPI Security Risk
- MITRE CWE-288 — Authentication Bypass Using Alternate Path
- BleepingComputer — HPE Aruba PAPI Critical Flaws
Why this verdict
- Management-plane restriction: PAPI on UDP 8211 is a management protocol confined to management VLANs in any properly segmented deployment. The CVSS AV:N score models internet-facing reachability, but real-world exposure is ~180 hosts globally. This single factor eliminates the majority of the unauthenticated-remote threat surface that justifies a 10.0.
- Unspecified preconditions: HPE explicitly states that "certain preconditions outside of the attacker's control must be met" — this contradicts the AC:L (low attack complexity) in the CVSS vector. Unknown preconditions represent real friction that the CVSS score ignores.
- No PoC or exploitation: Zero days after disclosure, there is no public exploit code, no Nuclei template, no Metasploit module, and no observed campaign. PAPI's proprietary binary protocol raises the skill bar for independent exploit development. Prior PAPI CVEs (CVE-2024-42509, CVE-2024-47460) also never saw public exploitation despite 9.8 scores.
- Role multiplier: AOS-10 is a canonical network infrastructure component — it runs on wireless APs, Gateways, and Mobility Conductors. On a standalone AP (typical role ~70% of installs), compromise yields traffic interception for that AP's clients and a lateral-movement pivot — blast radius is host-to-segment. On a Mobility Conductor or Gateway (high-value role ~15–20% of installs), compromise gives fleet-scale wireless infrastructure takeover: the attacker can push rogue configs to all managed APs, intercept RADIUS/EAP flows, and pivot to wired segments — blast radius is fleet-scale. Because this is a canonical network infrastructure component and the Conductor/Gateway scenario is fleet-scale with ≥10% of the installed base occupying that role, the verdict floor is HIGH.
- No active exploitation keeps this below CRITICAL floor: Despite the high-value-role floor, the combination of management-plane-only reachability, unspecified preconditions, and zero exploitation evidence means the real-world immediacy does not justify CRITICAL. The floor holds at HIGH.
Why not higher?
The floor for a canonical network infrastructure component with fleet-scale blast radius on Conductors/Gateways would be CRITICAL if the attack path were frictionless and internet-facing. However, PAPI's management-plane confinement means the attacker must already have internal network access or exploit a rare misconfiguration. HPE's own "preconditions outside attacker's control" language further limits the chain. No PoC or exploitation activity confirms that the weaponization barrier is non-trivial. CRITICAL would require either confirmed exploitation, a public PoC demonstrating trivial exploitation, or evidence of significant internet exposure — none of which exist.
Why not lower?
Dropping to MEDIUM would ignore the canonical high-value-role floor. AOS-10 Conductors and Gateways are fleet-critical infrastructure: compromising one gives an attacker control of every managed AP and visibility into all enterprise wireless traffic. Aruba holds 11–14% of the enterprise wireless market, meaning hundreds of thousands of devices are affected. The authentication bypass is unauthenticated — no credentials needed — and any attacker who has reached the management VLAN (common post-initial-access) gets one-packet RCE. The scope-change property is real: AP compromise changes the security boundary for all associated wireless clients. MEDIUM would be appropriate only if the blast radius were limited to a single low-value host, which it demonstrably is not.
Crowdsourced verification payload.
Run this script on any host with network access to the Aruba device management interface (e.g., a jump box on the management VLAN). Pass the target device IP as the first argument: bash check_cve_2026_79844.sh 10.1.1.100. Requires snmpget or curl for version detection — no elevated privileges needed on the auditor host. The script checks the AOS version via SNMP or HTTP and compares against known-vulnerable ranges.
#!/usr/bin/env bash
# check_cve_2026_79844.sh — Check HPE Aruba AOS-10 for CVE-2026-79844
# Usage: bash check_cve_2026_79844.sh <target_ip>
# Exit codes: 0=PATCHED, 1=VULNERABLE, 2=UNKNOWN
set -euo pipefail
TARGET="${1:-}"
if [[ -z "$TARGET" ]]; then
echo "Usage: $0 <target_ip>"
echo "UNKNOWN — no target specified"
exit 2
fi
echo "[*] Checking CVE-2026-79844 on $TARGET"
echo "[*] Attempting HTTPS version detection..."
# Try to grab version from the web UI login page or API
VERSION=""
if command -v curl &>/dev/null; then
# Aruba devices often expose version in the login page or REST API
RAW=$(curl -sk --connect-timeout 5 --max-time 10 "https://${TARGET}/rest/v1/configuration/object/version" 2>/dev/null || true)
if [[ -n "$RAW" ]]; then
VERSION=$(echo "$RAW" | grep -oP '10\.[0-9]+\.[0-9]+\.[0-9]+' | head -1 || true)
fi
# Fallback: try login page
if [[ -z "$VERSION" ]]; then
RAW=$(curl -sk --connect-timeout 5 --max-time 10 "https://${TARGET}/" 2>/dev/null || true)
VERSION=$(echo "$RAW" | grep -oP 'AOS[- ]?10[._]([0-9]+[._])+[0-9]+' | head -1 | sed 's/[_]/./' | grep -oP '10\.[0-9]+\.[0-9]+\.[0-9]+' || true)
fi
fi
# Fallback: try SNMP if available
if [[ -z "$VERSION" ]] && command -v snmpget &>/dev/null; then
echo "[*] HTTPS failed, trying SNMP..."
RAW=$(snmpget -v2c -c public -t 5 "$TARGET" 1.3.6.1.2.1.1.1.0 2>/dev/null || true)
VERSION=$(echo "$RAW" | grep -oP '10\.[0-9]+\.[0-9]+\.[0-9]+' | head -1 || true)
fi
if [[ -z "$VERSION" ]]; then
echo "UNKNOWN — could not detect AOS-10 version on $TARGET"
echo "[*] Verify manually: check 'show version' on the device CLI"
exit 2
fi
echo "[*] Detected AOS-10 version: $VERSION"
# Version comparison function
version_lte() {
[ "$(printf '%s\n' "$1" "$2" | sort -V | head -n1)" = "$1" ]
}
version_gte() {
[ "$(printf '%s\n' "$1" "$2" | sort -V | tail -n1)" = "$1" ]
}
# Check against vulnerable ranges
# Range 1: 10.3.0.0 - 10.4.1.13 (fixed: 10.4.1.14)
# Range 2: 10.6.0.0 - 10.7.2.6 (fixed: 10.7.2.7)
# Range 3: 10.8.0.0 - 10.8.1.1 (fixed: 10.8.1.2)
VULNERABLE=false
if version_gte "$VERSION" "10.3.0.0" && version_lte "$VERSION" "10.4.1.13"; then
echo "[!] Version $VERSION is in vulnerable range 10.3.0.0 - 10.4.1.13"
echo "[!] Upgrade to 10.4.1.14 or later"
VULNERABLE=true
elif version_gte "$VERSION" "10.6.0.0" && version_lte "$VERSION" "10.7.2.6"; then
echo "[!] Version $VERSION is in vulnerable range 10.6.0.0 - 10.7.2.6"
echo "[!] Upgrade to 10.7.2.7 or later"
VULNERABLE=true
elif version_gte "$VERSION" "10.8.0.0" && version_lte "$VERSION" "10.8.1.1"; then
echo "[!] Version $VERSION is in vulnerable range 10.8.0.0 - 10.8.1.1"
echo "[!] Upgrade to 10.8.1.2 or later"
VULNERABLE=true
elif version_gte "$VERSION" "10.5.0.0" && version_lte "$VERSION" "10.5.99.99"; then
echo "[!] Version $VERSION is in end-of-maintenance branch 10.5.x"
echo "[!] No patch available — migrate to a supported branch"
VULNERABLE=true
fi
if $VULNERABLE; then
echo "VULNERABLE — CVE-2026-79844 affects this device"
exit 1
else
echo "PATCHED — version $VERSION is not in a known vulnerable range"
exit 0
fi