GHSA-WG5R-WC3X-39VC
Vulnerability from github – Published: 2026-07-24 21:11 – Updated: 2026-08-03 20:44Summary
A pre-authentication remote code execution vulnerability affects OpenAM. The
remote authentication endpoint (/authservice, PLL) accepts an XML element
that names an arbitrary Java class, which the server then loads and
instantiates without validation. On a default configuration this is reachable
without authentication and allows an attacker to run code on the server.
Impact
Unauthenticated remote code execution / full server compromise on any OpenAM instance with default settings.
Affected
All releases up to and including 16.1.1 (the defect predates the Open Identity Platform fork).
Remediation
Upgrade to 16.1.2. The fix resolves the class named in a <CustomCallback>
element without running its static initialisers and rejects it unless it
implements DSAMECallbackInterface, and it constrains deserialisation of the
serialised Subject value to a class allowlist.
Interim mitigation
If you cannot upgrade immediately:
- Restrict or block external network access to
/authservice. This is the only reliable mitigation. - Optionally, block PLL requests carrying a
<CustomCallback className="...">element at the reverse proxy or WAF. That element is only produced for customDSAMECallbackInterfacecallbacks, so most deployments never send it — confirm against your own traffic before enforcing. - Enabling
sunRemoteAuthSecurityEnableddoes not mitigate this issue. The remote-auth security token is checked inAuthXMLHandler.processAuthXMLRequest, which runs only afterAuthXMLRequest.parseXMLhas already parsed the request and instantiated the class named in the<CustomCallback className="...">element. Do not rely on it as a substitute for upgrading or for network restriction.
Credit
Vulnerability discovered by Zhixi "Jace" Sun of ASM/VI at TikTok. Correction of the interim mitigation guidance contributed by @BarakSrour.
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 16.1.1"
},
"package": {
"ecosystem": "Maven",
"name": "org.openidentityplatform.openam:openam-core"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "16.1.2"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-62379"
],
"database_specific": {
"cwe_ids": [
"CWE-470",
"CWE-94"
],
"github_reviewed": true,
"github_reviewed_at": "2026-07-24T21:11:09Z",
"nvd_published_at": null,
"severity": "CRITICAL"
},
"details": "## Summary\nA pre-authentication remote code execution vulnerability affects OpenAM. The\nremote authentication endpoint (`/authservice`, PLL) accepts an XML element\nthat names an arbitrary Java class, which the server then loads and\ninstantiates without validation. On a default configuration this is reachable\n**without authentication** and allows an attacker to run code on the server.\n\n## Impact\nUnauthenticated remote code execution / full server compromise on any OpenAM\ninstance with default settings.\n\n## Affected\nAll releases up to and including 16.1.1 (the defect predates the Open Identity\nPlatform fork).\n\n## Remediation\nUpgrade to `16.1.2`. The fix resolves the class named in a `\u003cCustomCallback\u003e`\nelement without running its static initialisers and rejects it unless it\nimplements `DSAMECallbackInterface`, and it constrains deserialisation of the\nserialised `Subject` value to a class allowlist.\n\n## Interim mitigation\nIf you cannot upgrade immediately:\n\n- **Restrict or block external network access to `/authservice`.** This is the\n only reliable mitigation.\n- Optionally, **block PLL requests carrying a `\u003cCustomCallback className=\"...\"\u003e`\n element** at the reverse proxy or WAF. That element is only produced for custom\n `DSAMECallbackInterface` callbacks, so most deployments never send it \u2014 confirm\n against your own traffic before enforcing.\n- **Enabling `sunRemoteAuthSecurityEnabled` does *not* mitigate this issue.** The\n remote-auth security token is checked in `AuthXMLHandler.processAuthXMLRequest`,\n which runs only after `AuthXMLRequest.parseXML` has already parsed the request\n and instantiated the class named in the `\u003cCustomCallback className=\"...\"\u003e`\n element. Do not rely on it as a substitute for upgrading or for network\n restriction.\n\n## Credit\nVulnerability discovered by Zhixi \"Jace\" Sun of ASM/VI at TikTok.\nCorrection of the interim mitigation guidance contributed by @BarakSrour.",
"id": "GHSA-wg5r-wc3x-39vc",
"modified": "2026-08-03T20:44:06Z",
"published": "2026-07-24T21:11:09Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/OpenIdentityPlatform/OpenAM/security/advisories/GHSA-wg5r-wc3x-39vc"
},
{
"type": "WEB",
"url": "https://github.com/OpenIdentityPlatform/OpenAM/commit/edcf968cad91a78b932dba4ad559ef94cbf35f5a"
},
{
"type": "PACKAGE",
"url": "https://github.com/OpenIdentityPlatform/OpenAM"
},
{
"type": "WEB",
"url": "https://github.com/OpenIdentityPlatform/OpenAM/releases/tag/16.1.2"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
],
"summary": "OpenAM: Unauthenticated Remote Code Execution via Class.forName in AuthXMLUtils.createCustomCallback"
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.