CVE-2024-4839: LoLLMs CSRF — LOW

CISO Take

If your team uses lollms-webui for local LLM experimentation or internal AI tooling, this CSRF flaw lets attackers trick authenticated admins into silently installing AI services (XTTS, vLLM, Petals) via a crafted web page. The practical risk is unauthorized AI inference components being added to your environment without consent. Patch to the latest version and restrict the configuration interface to localhost or trusted networks only.

What is the risk?

CVSS 3.3 (Low) accurately reflects the limited blast radius: local attack vector, requires user interaction, no confidentiality impact. However, contextual risk is moderately elevated in AI development environments where lollms-webui runs with broad permissions and administrators may browse untrusted sites from the same machine. The ability to silently install AI services introduces unvetted compute-intensive components and potential supply chain exposure, which matters more in regulated or production-adjacent environments than the raw score suggests.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
LoLLMs	pip	—	No patch
81 Pushed 9d ago 33% patched ~1d to patch Full package profile →

Do you use LoLLMs? You're affected.

How severe is it?

CVSS 3.1

3.3 / 10

EPSS

0.2%

chance of exploitation in 30 days

Higher than 6% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

Exploit Available

Exploitation: MEDIUM

Sophistication

Trivial

Exploitation Confidence

medium

○ CISA SSVC: Public PoC

○ Public PoC indexed (trickest/cve)

Composite signal derived from CISA KEV, VulnCheck KEV, CISA SSVC, EPSS, Metasploit, Exploit-DB, trickest/cve, Nuclei templates, and inthewild.io exploitation reports.

What is the attack surface?

AV Local

AC Low

PR None

UI Required

S Unchanged

C None

I Low

A None

What should I do?

5 steps

Patch: Update lollms-webui to the latest available version where CSRF tokens have been applied to configuration endpoints.
Network segmentation: Bind the lollms-webui interface to 127.0.0.1 only; never expose admin endpoints to untrusted networks.
Browser hygiene: Warn users not to browse untrusted sites while lollms-webui is running in the same browser session.
Audit: Review currently installed services in existing deployments (check /etc/lollms or equivalent service directories) for unexpected additions.
Detection: Monitor for unexpected process launches (xtts-server, vllm, petals) and unusual outbound network connections from the host running lollms-webui.

What does CISA's SSVC say?

Decision Track*

Exploitation poc

Automatable No

Technical Impact partial

Source: CISA Vulnrichment (SSVC v2.0). Decision based on the CISA Coordinator decision tree.

How is it classified?

Supply Chain Auth Bypass Inference Framework AML.T0010.001 - AI Software AML.T0011 - User Execution AML.T0011.001 - Malicious Package AML.T0049 - Exploit Public-Facing Application

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act

Article 9 - Risk management system

ISO 42001

A.6.2.4 - AI system resources

NIST AI RMF

GOVERN 1.7 - Processes for identifying and managing AI risks

OWASP LLM Top 10

LLM05:2025 - Improper Output Handling / Supply Chain Vulnerabilities

Frequently Asked Questions

What is CVE-2024-4839?

If your team uses lollms-webui for local LLM experimentation or internal AI tooling, this CSRF flaw lets attackers trick authenticated admins into silently installing AI services (XTTS, vLLM, Petals) via a crafted web page. The practical risk is unauthorized AI inference components being added to your environment without consent. Patch to the latest version and restrict the configuration interface to localhost or trusted networks only.

Is CVE-2024-4839 actively exploited?

Proof-of-concept exploit code is publicly available for CVE-2024-4839, increasing the risk of exploitation.

How to fix CVE-2024-4839?

1. Patch: Update lollms-webui to the latest available version where CSRF tokens have been applied to configuration endpoints. 2. Network segmentation: Bind the lollms-webui interface to 127.0.0.1 only; never expose admin endpoints to untrusted networks. 3. Browser hygiene: Warn users not to browse untrusted sites while lollms-webui is running in the same browser session. 4. Audit: Review currently installed services in existing deployments (check /etc/lollms or equivalent service directories) for unexpected additions. 5. Detection: Monitor for unexpected process launches (xtts-server, vllm, petals) and unusual outbound network connections from the host running lollms-webui.

What systems are affected by CVE-2024-4839?

This vulnerability affects the following AI/ML architecture patterns: local AI development environments, LLM inference servers, AI model serving.

What is the CVSS score for CVE-2024-4839?

CVE-2024-4839 has a CVSS v3.1 base score of 3.3 (LOW). The EPSS exploitation probability is 0.16%.

What is the AI security impact?

Affected AI Architectures

local AI development environmentsLLM inference serversAI model serving

MITRE ATLAS Techniques

AML.T0010.001 AI Software

AML.T0011 User Execution

AML.T0011.001 Malicious Package

AML.T0049 Exploit Public-Facing Application

Compliance Controls Affected

EU AI Act: Article 9

ISO 42001: A.6.2.4

NIST AI RMF: GOVERN 1.7

OWASP LLM Top 10: LLM05:2025

What are the technical details?

Original Advisory

A Cross-Site Request Forgery (CSRF) vulnerability exists in the 'Servers Configurations' function of the parisneo/lollms-webui, versions 9.6 to the latest. The affected functions include Elastic search Service (under construction), XTTS service, Petals service, vLLM service, and Motion Ctrl service, which lack CSRF protection. This vulnerability allows attackers to deceive users into unwittingly installing the XTTS service among other packages by submitting a malicious installation request. Successful exploitation results in attackers tricking users into performing actions without their consent.

Exploitation Scenario

An attacker targeting a data science team creates a malicious website containing a hidden HTML form that auto-submits a POST request to http://localhost:9600/install_service with payload {"service": "xtts"}. When a developer who has lollms-webui open in the same browser visits a phishing link (e.g., disguised as a research paper or a meme link on Slack), their browser sends the request with active session cookies. lollms-webui, lacking CSRF validation, processes the request and begins downloading and installing the XTTS service in the background. The developer sees no prompt; the attacker has now introduced an AI speech-synthesis service running locally with whatever network access lollms-webui has.

Weaknesses (CWE)

CWE-352 Cross-Site Request Forgery (CSRF)

CWE-352 — Cross-Site Request Forgery (CSRF): The web application does not, or cannot, sufficiently verify whether a request was intentionally provided by the user who sent the request, which could have originated from an unauthorized actor.

[Architecture and Design] Use a vetted library or framework that does not allow this weakness to occur or provides constructs that make this weakness easier to avoid [REF-1482]. For example, use anti-CSRF packages such as the OWASP CSRFGuard. [REF-330] Another example is the ESAPI Session Management control, which includes a component for CSRF. [REF-45]
[Implementation] Ensure that the application is free of cross-site scripting issues (CWE-79), because most CSRF defenses can be bypassed using attacker-controlled script.

Source: MITRE CWE corpus.