As part of its Third-Party Vulnerability Disclosure (3PVD) program, Google published a proof-of-concept (PoC) exploit for a high-severity command injection vulnerability in Palo Alto Networks’ PAN-OS OpenConfig plugin.
Tracked as CVE-2025-0110, the flaw allows authenticated administrators to execute arbitrary commands on firewalls via manipulated gNMI requests, escalating privileges to root access.
The disclosure follows Palo Alto Networks’ February 2025 patch release and highlights growing concerns about firewall exploitation chains in critical infrastructure.
CVE-2025-0110 resides in the PAN-OS OpenConfig plugin, which facilitates network device configuration via the gNMI protocol.
Attackers exploiting this flaw can bypass security restrictions by injecting malicious commands into the type parameter of an XPATH query during syslog retrieval. For example, the PoC demonstrates embedding $(echo system > file1; cat file1) into the query to execute bash commands.
bash./gnmic -a <IP>:9339 -u admin --password=<PASSWORD> --skip-verify \
--path 'pan-logging:/pan/logging/query/custom[type=$(echo system > file1; cat file1)]' Successful exploitation allows attackers to reconfigure firewalls, exfiltrate sensitive data, or deploy persistent backdoors like the UPSTYLE malware observed in prior PAN-OS campaigns.
While CVE-2025-0110 requires authentication, Google’s researchers emphasize its danger when combined with CVE-2025-0108, an authentication bypass flaw patched earlier this month. Threat actors could chain these vulnerabilities to:
Palo Alto Networks confirmed active exploitation of this chained attack vector, with GreyNoise observing 26 malicious IPs targeting exposed management interfaces.
Palo Alto Networks released fixed OpenConfig plugin versions (≥2.1.2) on February 12, 2025, urging customers to:
Google’s disclosure aligns with its 90-day vulnerability disclosure policy, noting that patches were available prior to publication. However, Shadowserver Foundation reports over 3,500 internet-exposed PAN-OS interfaces remain unsecured as of February 21.
Free Webinar: Better SOC with Interactive Malware Sandbox for Incident Response and Threat Hunting – Register Here
Hackers are actively probing AI systems, turning exposed gateways and agent tools into routes for…
Hackers are making some phishing pages harder to track by changing the code delivered to…
A cyber incident reportedly forced a British power plant to halt operations for about four…
Russian hackers have used a new backdoor called HOOKEDGE to target defense manufacturers, government bodies,…
TITAN ransomware is pairing file encryption with an ambitious claim: artificial intelligence that can sort…
A fake student resume is being used to place a remote-access tool on researchers’ Windows…