/

CVE-2026-82860

Severity:

CRITICAL RISK

(9.8)

Improper Privilege Management (CWE-269) in Hulumi Policies

CVE-2026-82860

Deferred

Vulnerability type

Category:

Improper Privilege Management

CWE:

CWE-269

Impact:

Authorization Bypass and Privilege Escalation

Severity:

CRITICAL RISK

(9.8)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

@hulumi/policies versions before 1.3.2 do not completely inspect inline and attached IAM policy evidence when enforcing the administrator-policy guardrail. An attacker who can supply or influence policy definitions may craft admin-equivalent policy paths that bypass evaluation controls, potentially gaining excessive privileges.

Affected context

The vulnerability occurs in deployments that use the affected @hulumi/policies administrator-policy guardrail to assess IAM policies. Risk is present when policy evidence can be supplied, modified, or attached through an exposed infrastructure-management workflow and the evaluation logic does not fully account for all inline and attached policy relationships.

Exploit preconditions

  • Attack surface: IAM policy evaluation within the @hulumi/policies administrator-policy guardrail
  • Required access: Ability to create, modify, attach, or otherwise influence policy definitions or their evaluated evidence
  • Trigger point: Guardrail analysis of inline and attached IAM policies
  • Why exploitation is possible: Incomplete inspection can allow admin-equivalent policy paths to evade authorization checks

Why this happens

The underlying bug is incomplete privilege analysis during IAM policy evaluation. The guardrail does not fully inspect all relevant inline and attached policy evidence, allowing effective permissions to differ from the permissions recognized by the control. This maps to CWE-269, Improper Privilege Management, because authorization decisions may be made without accurately accounting for the privileges granted through indirect or alternative policy paths.

Potential impact

  • An attacker may bypass administrator-policy guardrails
  • Unauthorized users or roles could gain elevated IAM permissions
  • Attackers could modify infrastructure, identities, or access policies
  • Cloud resources or sensitive data could be exposed or disrupted
  • Compromised privileges could enable lateral movement across connected services

General mitigation guidance

Mitigation generally focuses on ensuring that authorization analysis reflects effective permissions rather than only directly visible policy statements.

Common defensive patterns include:

  • Use policy evaluation logic that considers inline, attached, inherited, and indirectly effective permissions
  • Apply secure-by-default deny behavior when policy evidence is incomplete or ambiguous
  • Reduce input trust by validating policy sources, relationships, and attachment metadata
  • Enforce least privilege for identities that can create or modify IAM policies
  • Separate policy authoring from policy approval and deployment workflows
  • Add independent authorization checks for high-impact administrative actions

How teams detect issues like this

Organizations can identify similar weaknesses through:

  • Static analysis for policy evaluators that inspect only direct or inline statements
  • Code review checks confirming that attached and inherited policy relationships are analyzed
  • Negative tests covering equivalent privilege paths, wildcard permissions, and indirect attachments
  • Policy-as-code enforcement that compares declared permissions with effective permissions
  • Audit-log monitoring for unexpected policy creation, attachment, or privilege changes
  • Automated security review tools that flag incomplete IAM graph traversal or fail-open evaluation behavior

Sources

SourceLink
github.comgithub.com/kerberosmansour/hulumi/security/advisories/GHSA-4xrh-5m3m-328w
vulncheck.comvulncheck.com/advisories/hulumi-policies-before-1.3.2-admin-policy-bypass

Frequently asked

questions

Exploitation depends on whether an attacker can influence IAM policy definitions, attachments, or evidence processed by the administrator-policy guardrail. The issue is particularly relevant in automated infrastructure and cloud-identity workflows.

Typically, yes. An attacker generally needs permission to create, modify, attach, or submit policy material that reaches the vulnerable evaluation path. The required permissions depend on the deployment architecture.

Use authorization analysis that calculates effective permissions across inline and attached policies, and fail closed when policy relationships or evidence cannot be fully evaluated.

Apply least privilege, review all policy attachment paths, test equivalent privilege scenarios, and enforce independent policy-as-code and audit controls for administrative permissions.

Find out what's already
exploitable in your codebase.