Vulnerability Management

Palo Alto GlobalProtect Zero-Day: Response Timeline

CVE-2024-3400 hit GlobalProtect with pre-auth RCE and ongoing exploitation. Here is the response timeline, the UPSTYLE tradecraft, and what worked.

Shadab Khan
Security Engineer
5 min read

On April 12, 2024, Palo Alto Networks disclosed CVE-2024-3400, a pre-authentication command injection in the GlobalProtect feature of PAN-OS that was already being exploited in the wild. The vulnerability carried a CVSS of 10.0. Volexity — the research team credited with initial discovery on April 10 while responding to a customer incident — published a detailed report naming the threat actor UTA0218 and documenting a custom Python backdoor dubbed UPSTYLE. CISA added the CVE to KEV on April 12 and issued a three-week remediation deadline. By April 14, mass exploitation was underway from multiple actors copying the tradecraft Volexity had reverse-engineered. This post walks the timeline hour by hour where it mattered, names the technical specifics, and pulls out the moves that reduced incident-response time at the organizations that fared best.

What is CVE-2024-3400 technically?

CVE-2024-3400 is a command injection in the GlobalProtect portal and gateway features that an unauthenticated remote attacker can exploit via a crafted session cookie, yielding code execution as root on PAN-OS. The vulnerable code path processes a telemetry-related session ID and passes user-controlled input to a shell context without sanitization. Affected configurations required both GlobalProtect and device telemetry to be enabled — a non-default but extremely common combination, especially on devices that had been upgraded through multiple PAN-OS minor versions. The bug reached into PAN-OS 10.2, 11.0, and 11.1, with fixes published in 10.2.9-h1, 11.0.4-h1, and 11.1.2-h3 and subsequent hotfixes.

What did UTA0218 actually do once inside?

UTA0218 dropped the UPSTYLE backdoor, created reverse shells, harvested firewall configuration and credentials, and then pivoted laterally into the internal network the firewall protected. Volexity's write-up describes UPSTYLE as a Python script persisted through a crontab entry that poked a file path for commands to execute, giving the actor an interactive C2 channel without opening obvious listening ports. The operator pattern was consistent: exfiltrate running-config.xml and credential stores, enumerate the internal environment via the firewall's own management interfaces, and then move to cloud administrative consoles when the configs revealed SSO or federation relationships. Evidence of targeted intent — the actor focused on specific high-value organizations rather than mass spray — suggested nation-state involvement, though no public attribution to a named country was confirmed at time of writing.

How did the response timeline actually unfold?

Volexity detected exploitation at a customer on April 10, reported to Palo Alto the same day, and Palo Alto issued the advisory and interim mitigations on April 12. The order mattered. On April 12, Palo Alto recommended disabling device telemetry as a workaround, which effectively closed the vulnerable code path without a firmware upgrade. On April 14, Palo Alto reversed position after realizing the telemetry workaround was insufficient in some cases, and urged customers to apply the hotfix and upgrade. CISA added to KEV the same day as disclosure. The mid-course reversal cost hours of IR time at organizations that had treated the telemetry workaround as complete, and it highlighted how often initial mitigations do not survive the first week.

Who got exploited and who avoided it?

Organizations with public-facing GlobalProtect portals on PAN-OS 10.2, 11.0, or 11.1 and with device telemetry enabled were the primary victims, while those running older PAN-OS, those fronting GlobalProtect with an IP allowlist, and those with telemetry disabled largely avoided exploitation. Volexity's reporting and subsequent incident response work indicated victims across government, education, and enterprise verticals. The organizations that avoided UPSTYLE generally did one of three things: they had upstream IP reputation filtering that dropped scanner and exploit traffic; they had disabled telemetry months earlier for bandwidth reasons; or they ran GlobalProtect behind a VPN-only exposure enforced at the perimeter rather than directly on the public internet.

What did effective IR look like during the first 72 hours?

Effective incident response meant upgrading PAN-OS and rotating every credential and key that touched the firewall rather than trusting the workaround. The organizations that fared best executed a tight playbook: apply hotfix or upgrade first, generate a support file (tech support file) and review it for UPSTYLE artifacts, rotate all admin credentials, rotate any SAML signing certificates and IPSec PSKs stored in running-config.xml, force a re-auth on all GlobalProtect users, and audit the six months of configuration change history for unauthorized modifications. Hunting the UPSTYLE backdoor itself is straightforward once you have shell access to the device, but most organizations did not have shell access and had to rely on Palo Alto's own forensic guidance through TAC.

What is the durable lesson for edge-device operations?

The durable lesson is to treat every edge appliance as a full compromise candidate on every CVE cycle, not just as a thing to patch. PAN-OS CVE-2024-3400 followed the same playbook as Ivanti Connect Secure CVE-2024-21887, FortiOS CVE-2024-21762, and ConnectWise ScreenConnect CVE-2024-1709 — pre-auth RCE on an internet-facing appliance, used by at least one capable actor before disclosure, then mass-exploited within 48 hours of public details. The edge is the new software supply chain front: one appliance vendor's bug grants a foothold in thousands of downstream networks.

# Minimum triage on a PAN-OS device post-patch
show system info | match sw-version
show running config | match telemetry
request tech-support dump
# Export and inspect for UPSTYLE artefacts in /var/log/pan/

How Safeguard Helps

Safeguard inventories PAN-OS firewalls as supply chain assets and cross-checks firmware versions against the CVE-2024-3400 advisory, so reachability analysis shows which internal systems are downstream of a potentially compromised GlobalProtect portal. Griffin AI matches Volexity's UTA0218 indicators against your flow data and surfaces likely victim configurations before the wider industry writes them up. SBOMs record every firmware build across PAN-OS 10.2, 11.0, and 11.1, and policy gates block new deployments that depend on components with unresolved KEV advisories. TPRM assessments flag vendors and managed network providers operating affected appliances so your contract language can demand remediation evidence before their exposure becomes yours.

Never miss an update

Weekly insights on software supply chain security, delivered to your inbox.