CVE-2026-32617
Published: 16 March 2026
Summary
CVE-2026-32617 is a high-severity Permissive Cross-domain Security Policy with Untrusted Domains (CWE-942) vulnerability in Mintplexlabs Anythingllm. Its CVSS base score is 7.1 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique External Remote Services (T1133); ranked at the 9.9th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.
This vulnerability is AI-related — categorised as LLM Application Platforms; in the Supply Chain and Deployment risk domain.
The strongest mitigations our analysis identified are NIST 800-53 AC-3 (Access Enforcement) and IA-2 (Identification and Authentication (Organizational Users)).
Deeper analysis
CVE-2026-32617 is a vulnerability in AnythingLLM, an application that converts content into context for large language models (LLMs) during chats. It affects versions 1.11.1 and earlier, specifically default installations where no password or API key is configured. In these setups, all HTTP endpoints and the agent WebSocket lack authentication, while the server's CORS policy permits requests from any origin. The AnythingLLM Desktop edition binds to 127.0.0.1 (loopback) by default. The issue is classified under CWE-942 (Permissive Cross-domain Policy with Untrusted Domains) and CWE-1188 (Implementation of a web server with an insecure default configuration), with a CVSS v3.1 base score of 7.1 (AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:L).
Exploitation requires an attacker to be on the same local network (LAN), as modern browsers (Chrome, Edge, Firefox) enforce Private Network Access (PNA), blocking public websites from accessing local IP addresses. An attacker with network access but no privileges can leverage the lack of authentication and permissive CORS, though it demands high attack complexity and user interaction. Successful exploitation enables high confidentiality and integrity impacts, such as unauthorized access to or modification of LLM context data, with low availability impact.
The GitHub security advisory provides details on mitigation: https://github.com/Mintplex-Labs/anything-llm/security/advisories/GHSA-24qj-pw4h-3jmm. Published on 2026-03-16, no real-world exploitation or additional AI/ML-specific context is noted in available information.
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2026-12103
Vulnerability details
AnythingLLM is an application that turns pieces of content into context that any LLM can use as references during chatting. In 1.11.1 and earlier, On default installations where no password or API key has been configured, all HTTP endpoints and…
more
the agent WebSocket lack authentication, and the server's CORS policy accepts any origin. AnythingLLM Desktop binds to 127.0.0.1 (loopback) by default. Modern browsers (Chrome, Edge, Firefox) implement Private Network Access (PNA). This explicitly blocks public websites from making requests to local IP addresses. Exploitation is only viable from within the same local network (LAN) due to browser-level blocking of public-to-private requests.
- CWE(s)
AI Security AnalysisAI
- AI Category
- LLM Application Platforms
- Risk Domain
- Supply Chain and Deployment
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- Matched keywords: anythingllm, llm
Related Threats
MITRE ATT&CK Enterprise TechniquesAI
Why these techniques?
Insecure default configuration (no auth on all HTTP/WebSocket endpoints + permissive CORS) directly enables initial access by exploiting the exposed AnythingLLM service or application without credentials.
CVEs Like This One
Affected Assets
Mitigating Controls
Mitigating Controls (NIST 800-53 r5) AI
Directly enforces authentication and authorization checks on all HTTP and WebSocket endpoints before any access is granted, eliminating the unauthenticated exposure described in the CVE.
Requires unique identification and authentication of users prior to allowing access to the application interfaces, directly countering the default no-password/API-key configuration.
Mandates applying restrictive configuration settings (including enabling authentication and tightening CORS) instead of the insecure defaults that leave endpoints open.