What is the CVSS score of CVE-2026-32900?

CVE-2026-32900 has a CVSS 3.1 base score of 6.4 (Medium). CVSS vector: CVSS:3.1/AV:N/AC:H/PR:H/UI:R/S:U/C:H/I:H/A:H.

Which versions of Openclaw are affected by CVE-2026-32900?

Organizations using OpenClaw before 2026.2.22 should be aware of moderate risk from CVE-2026-32900, a incorrect authorization vulnerability rated CVSS 6.4.. A patch is available.

How to fix or mitigate CVE-2026-32900?

Within 30 days: Identify affected systems and apply vendor patches as part of regular patch cycle. Vendor patch is available.

Openclaw CVE-2026-32900

EUVD-2026-14578 MEDIUM

Incorrect Authorization (CWE-863)

2026-03-23 VulnCheck

Authentication Bypass Openclaw

6.4

CVSS 3.1 · NVD

Severity by source

NVD PRIMARY

6.4 MEDIUM

AV:N/AC:H/PR:H/UI:R/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:H/PR:H/UI:R/S:U/C:H/I:H/A:H

Attack Vector

Network

Attack Complexity

High

Privileges Required

High

User Interaction

Required

Scope

Unchanged

Confidentiality

High

Integrity

High

Availability

High

Lifecycle Timeline

EUVD ID Assigned

Mar 23, 2026 - 22:00 euvd

EUVD-2026-14578

Analysis Generated

Mar 23, 2026 - 22:00 vuln.today

Patch released

Mar 23, 2026 - 22:00 nvd

Patch available

CVE Published

Mar 23, 2026 - 21:36 nvd

MEDIUM 6.4

DescriptionCVE.org

OpenClaw before 2026.2.22 contains an authorization bypass vulnerability in allowlist mode where allow-always persistence at wrapper-level enables approval-bypass execution of different payloads. Attackers can approve benign wrapped system.run commands to broaden trust boundaries and execute arbitrary commands without subsequent approval on gateway and node-host execution flows.

AnalysisAI

OpenClaw before version 2026.2.22 contains an authorization bypass vulnerability in allowlist mode that allows attackers with high privileges to approve benign wrapped system.run commands and subsequently execute arbitrary commands without requiring additional approval on gateway and node-host execution flows. This vulnerability exploits allow-always persistence at the wrapper level to broaden trust boundaries beyond the initial approval scope. The vulnerability has a CVSS score of 6.4 with high impact on confidentiality, integrity, and availability, though exploitation requires high privilege level and user interaction.

Technical ContextAI

The vulnerability exists in OpenClaw's authorization framework, specifically within the allowlist mode security model where wrapper-level approval mechanisms are implemented. According to CWE-863 (Incorrect Authorization), the root cause is improper authorization enforcement that fails to adequately validate subsequent command execution against the original approval scope. The affected product OpenClaw (CPE: cpe:2.3:a:openclaw:openclaw:*:*:*:*:*:*:*:*) implements a gateway and node-host execution flow architecture where approval decisions should be persistent but scoped to specific commands. The allow-always persistence mechanism incorrectly extends trust from benign wrapped system.run commands to arbitrary payloads, effectively converting a single-command approval into a blanket authorization for the trust boundary. This is particularly dangerous in multi-stage execution scenarios where attackers can chain previously approved operations to execute malicious code that bypasses intended approval checkpoints.

RemediationAI

Upgrade OpenClaw to version 2026.2.22 or later, which includes the authorization bypass fix documented in commit 24c954d972400f508814532dea0e4dcb38418bb0 (available at https://github.com/openclaw/openclaw/commit/24c954d972400f508814532dea0e4dcb38418bb0). Until patching is completed, implement compensating controls by restricting high-privilege account access to OpenClaw allowlist mode approval functions, enforcing multi-factor authentication for accounts with approval authority, and conducting immediate audit of all previously approved system.run commands to identify any that may have been exploited for broader command execution. Additionally, monitor gateway and node-host execution flows for commands that do not match approved allowlist entries and implement network-level segmentation to restrict OpenClaw component communication to trusted internal networks only.