CVE-2026-44843
LangChain: Unsafe deserialization of attacker-controlled LangChain objects through overly broad `load()` allowlists
Description
LangChain is a framework for building agents and LLM-powered applications. Prior to 0.3.85 and 1.3.3, LangChain contains older runtime code paths that deserialize run inputs, run outputs, or other application-controlled payloads using overly broad object allowlists. These paths may call load() with allowed_objects="all". This does not enable arbitrary Python object deserialization, but it does allow any trusted LangChain-serializable object to be revived, which is broader than these runtime paths require. As a result, attacker-supplied LangChain serialized constructor dictionaries may cause trusted runtime paths to instantiate classes with untrusted constructor arguments. This vulnerability is fixed in 0.3.85 and 1.3.3.
INFO
Published Date :
May 26, 2026, 9:16 p.m.
Last Modified :
July 24, 2026, 11:10 a.m.
Remotely Exploit :
Yes !
Source :
[email protected]
Affected Products
The following products are affected by CVE-2026-44843
vulnerability.
Even if cvefeed.io is aware of the exact versions of the
products
that
are
affected, the information is not represented in the table below.
CVSS Scores
| Score | Version | Severity | Vector | Exploitability Score | Impact Score | Source |
|---|---|---|---|---|---|---|
| CVSS | 134c704f-9b21-4f2e-91b3-4a467353bcc0 | |||||
| CVSS 3.1 | HIGH | [email protected] |
Solution
- Update LangChain to version 0.3.85 or later.
- Update LangChain to version 1.3.3 or later.
Public PoC/Exploit Available at Github
CVE-2026-44843 has a 2 public
PoC/Exploit available at Github.
Go to the Public Exploits tab to see the list.
References to Advisories, Solutions, and Tools
Here, you will find a curated list of external links that provide in-depth
information, practical solutions, and valuable tools related to
CVE-2026-44843.
| URL | Resource |
|---|---|
| https://github.com/langchain-ai/langchain/security/advisories/GHSA-pjwx-r37v-7724 | Mitigation Vendor Advisory |
CWE - Common Weakness Enumeration
While CVE identifies
specific instances of vulnerabilities, CWE categorizes the common flaws or
weaknesses that can lead to vulnerabilities. CVE-2026-44843 is
associated with the following CWEs:
Common Attack Pattern Enumeration and Classification (CAPEC)
Common Attack Pattern Enumeration and Classification
(CAPEC)
stores attack patterns, which are descriptions of the common attributes and
approaches employed by adversaries to exploit the CVE-2026-44843
weaknesses.
We scan GitHub repositories to detect new proof-of-concept exploits. Following list is a collection of public exploits and proof-of-concepts, which have been published on GitHub (sorted by the most recently updated).
A CLI tool that takes a task description, automatically generates 3 strategically distinct prompts using an LLM, runs all 3 against test cases in parallel, and scores each output on Accuracy, Brevity, and Helpfulness using a calibrated judge agent.
Python
Plug-and-Play AI Knowledge Chatbot Backend with Memory + RAG (LangGraph + Redis + ChromaDB)
Python Dockerfile HTML TypeScript CSS Java JavaScript Shell
Results are limited to the first 15 repositories due to potential performance issues.
The following list is the news that have been mention
CVE-2026-44843 vulnerability anywhere in the article.
The following table lists the changes that have been made to the
CVE-2026-44843 vulnerability over time.
Vulnerability history details can be useful for understanding the evolution of a vulnerability, and for identifying the most recent changes that may impact the vulnerability's severity, exploitability, or other characteristics.
-
CVE Translated by [email protected]
Jul. 24, 2026
Action Type Old Value New Value Changed Translation Title: LangChain de langchain-ai, Description: LangChain es un framework para construir agentes y aplicaciones impulsadas por LLM. Antes de 0.3.85 y 1.3.3, LangChain contiene rutas de código de tiempo de ejecución más antiguas que deserializan entradas de ejecución, salidas de ejecución u otras cargas útiles controladas por la aplicación utilizando listas de permitidos de objetos excesivamente amplias. Estas rutas pueden llamar a load() con allowed_objects='all'. Esto no habilita la deserialización arbitraria de objetos Python, pero sí permite que cualquier objeto LangChain serializable de confianza sea revivido, lo cual es más amplio de lo que requieren estas rutas de tiempo de ejecución. Como resultado, los diccionarios de constructor serializados de LangChain suministrados por el atacante pueden hacer que las rutas de tiempo de ejecución de confianza instancien clases con argumentos de constructor no confiables. Esta vulnerabilidad está corregida en 0.3.85 y 1.3.3. Title: LangChain de langchain-ai, Description: LangChain es un framework para construir agentes y aplicaciones impulsadas por LLM. Antes de 0.3.85 y 1.3.3, LangChain contiene rutas de código de tiempo de ejecución más antiguas que deserializan entradas de ejecución, salidas de ejecución u otras cargas útiles controladas por la aplicación utilizando listas de permitidos de objetos excesivamente amplias. Estas rutas pueden llamar a load() con allowed_objects='all'. Esto no habilita la deserialización arbitraria de objetos Python, pero sí permite que cualquier objeto LangChain serializable de confianza sea revivido, lo cual es más amplio de lo que requieren estas rutas de tiempo de ejecución. Como resultado, los diccionarios de constructor serializados de LangChain suministrados por el atacante pueden hacer que las rutas de tiempo de ejecución de confianza instancien clases con argumentos de constructor no confiables. Esta vulnerabilidad está corregida en 0.3.85 y 1.3.3. -
CVE Translated by [email protected]
Jul. 23, 2026
Action Type Old Value New Value Added Translation Title: LangChain de langchain-ai, Description: LangChain es un framework para construir agentes y aplicaciones impulsadas por LLM. Antes de 0.3.85 y 1.3.3, LangChain contiene rutas de código de tiempo de ejecución más antiguas que deserializan entradas de ejecución, salidas de ejecución u otras cargas útiles controladas por la aplicación utilizando listas de permitidos de objetos excesivamente amplias. Estas rutas pueden llamar a load() con allowed_objects='all'. Esto no habilita la deserialización arbitraria de objetos Python, pero sí permite que cualquier objeto LangChain serializable de confianza sea revivido, lo cual es más amplio de lo que requieren estas rutas de tiempo de ejecución. Como resultado, los diccionarios de constructor serializados de LangChain suministrados por el atacante pueden hacer que las rutas de tiempo de ejecución de confianza instancien clases con argumentos de constructor no confiables. Esta vulnerabilidad está corregida en 0.3.85 y 1.3.3. -
CVE Modified by [email protected]
Jun. 17, 2026
Action Type Old Value New Value Added Affected [{'vendor': 'langchain-ai', 'product': 'langchain', 'versions': [{'status': 'affected', 'version': '< 0.3.85'}, {'status': 'affected', 'version': '>= 1.0.0a1, < 1.3.3'}]}] -
CVE Modified by 134c704f-9b21-4f2e-91b3-4a467353bcc0
Jun. 17, 2026
Action Type Old Value New Value Added SSVC {'id': 'CVE-2026-44843', 'role': 'CISA Coordinator', 'options': [{'exploitation': 'none'}, {'automatable': 'yes'}, {'technicalImpact': 'partial'}], 'version': '2.0.3', 'timestamp': '2026-05-27T14:06:50.805026Z'} -
Initial Analysis by [email protected]
May. 29, 2026
Action Type Old Value New Value Added CPE Configuration OR *cpe:2.3:a:langchain:langchain:*:*:*:*:*:*:*:* versions up to (excluding) 0.3.85 *cpe:2.3:a:langchain:langchain:*:*:*:*:*:*:*:* versions from (including) 1.0.0 up to (excluding) 1.3.3 Added Reference Type GitHub, Inc.: https://github.com/langchain-ai/langchain/security/advisories/GHSA-pjwx-r37v-7724 Types: Mitigation, Vendor Advisory -
New CVE Received by [email protected]
May. 26, 2026
Action Type Old Value New Value Added Description LangChain is a framework for building agents and LLM-powered applications. Prior to 0.3.85 and 1.3.3, LangChain contains older runtime code paths that deserialize run inputs, run outputs, or other application-controlled payloads using overly broad object allowlists. These paths may call load() with allowed_objects="all". This does not enable arbitrary Python object deserialization, but it does allow any trusted LangChain-serializable object to be revived, which is broader than these runtime paths require. As a result, attacker-supplied LangChain serialized constructor dictionaries may cause trusted runtime paths to instantiate classes with untrusted constructor arguments. This vulnerability is fixed in 0.3.85 and 1.3.3. Added CVSS V3.1 AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:L/A:N Added CWE CWE-502 Added Reference https://github.com/langchain-ai/langchain/security/advisories/GHSA-pjwx-r37v-7724