Severity by source
AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H
Primary rating from NVD · only source for this CVE.
CVSS VectorNVD
CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:U/C:H/I:H/A:H
Lifecycle Timeline
10Blast Radius
ecosystem impact- 10 pypi packages depend on apache-airflow (1 direct, 9 indirect)
- 10 pypi packages depend on apache-airflow-core (2 direct, 8 indirect)
Ecosystem-wide dependent count for version 3.2.0 and other introduced versions.
DescriptionCVE.org
Dag Authors, who normally should not be able to execute code in the webserver context could craft XCom payload causing the webserver to execute arbitrary code. Since Dag Authors are already highly trusted, severity of this issue is Low.
Users are recommended to upgrade to Apache Airflow 3.2.0, which fixes the issue.
AnalysisAI
Deserialization vulnerability in Apache Airflow webserver (all versions before 3.2.0) allows network-accessible attackers to execute arbitrary code by injecting malicious XCom payloads, despite vendor-assigned Low severity due to the trusted Dag Author threat model. CVSS 9.8 Critical rating reflects unauthenticated network-based RCE capability (AV:N/PR:N), contradicting the description's trust assumption. EPSS 0.07% (22nd percentile) suggests low immediate exploitation likelihood. No active exploitation confirmed; vendor patch available in version 3.2.0 with public GitHub PR.
Technical ContextAI
Apache Airflow is a workflow orchestration platform where Dag Authors define task workflows. XCom (cross-communication) is Airflow's mechanism for passing data between tasks, traditionally stored and retrieved via serialization. This vulnerability stems from CWE-502 (Deserialization of Untrusted Data) where the webserver component improperly deserializes XCom payloads without sufficient validation. Attackers can craft malicious serialized objects that, when deserialized by the webserver, execute arbitrary Python code in the webserver's process context. The affected product CPE (cpe:2.3:a:apache_software_foundation:apache_airflow) covers all Airflow versions prior to 3.2.0. Python deserialization flaws typically involve pickle or similar unsafe serialization libraries that can instantiate arbitrary classes and invoke code during object reconstruction.
RemediationAI
Upgrade to Apache Airflow 3.2.0 or later, which addresses the unsafe XCom deserialization flaw (vendor advisory: https://lists.apache.org/thread/6whgpkqbh12rvpfmvcg8b0vwlv4hq3po, fix PR: https://github.com/apache/airflow/pull/61641). If immediate upgrade is not feasible, implement strict role-based access controls to ensure only highly trusted administrators hold Dag Author privileges, and audit existing Dag Authors for insider risk since the attack requires this role. Consider network segmentation to restrict webserver access to authenticated internal users only, reducing the practical attack surface despite the CVSS network vector rating. Deploy web application firewall rules to inspect and block suspicious XCom payloads if feasible, though deserialization attacks are difficult to detect at the network layer. Monitor webserver logs for unusual XCom operations or unexpected process executions. Note that restricting Dag Author privileges may impact legitimate workflow development workflows, requiring process adjustments.
Same weakness CWE-502 – Deserialization of Untrusted Data
View allSame technique Deserialization
View allShare
External POC / Exploit Code
Leaving vuln.today
EUVD-2026-23658
GHSA-6ffj-2wg2-w45j