High severity · published Dec 23, 2025
CVE-2025-68664
LangChain serialization injection vulnerability enables secret extraction in dumps/loads APIs
8.2
CVSS
42.9%
EPSS
No
On KEV
Description
LangChain is a framework for building agents and LLM-powered applications. Prior to versions 0.3.81 and 1.2.5, a serialization injection vulnerability exists in LangChain's dumps() and dumpd() functions. The functions do not escape dictionaries with 'lc' keys when serializing free-form dictionaries. The 'lc' key is used internally by LangChain to mark serialized objects. When user-controlled data contains this key structure, it is treated as a legitimate LangChain object during deserialization rather than plain user data. This issue has been patched in versions 0.3.81 and 1.2.5.
CVSS vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:L/A:N
Attack vector
Network
Complexity
Low
Privileges
None
User interaction
None
Confidentiality
High
Integrity
Low
Availability
None
Scope
Unchanged
Exploitation
Is it being used?
Exploit prediction (EPSS)42.9% · higher than 99% of CVEs
Exploited in the wildNo evidence yet
CISA KEVNot listed
Ransomware campaign useNot known
Affected products
Which versions
LA
langchain · langchain-ai
>= 1.0.0, < 1.2.5; < 0.3.81 · source: vendor
LC
Langchain Core · Langchain
before 0.3.81; 1.0.0 before 1.2.5 · source: NVD
References
NVD
https://nvd.nist.gov/vuln/detail/CVE-2025-68664
CVE.org
https://www.cve.org/CVERecord?id=CVE-2025-68664
Patch
https://github.com/langchain-ai/langchain/commit/5ec0fa69de31bbe3d76e4cf9cd65a6accb8466c8
Patch
https://github.com/langchain-ai/langchain/commit/d9ec4c5cc78960abd37da79b0250f5642e6f0ce6
Issue Tracking
https://github.com/langchain-ai/langchain/pull/34455
Issue Tracking
https://github.com/langchain-ai/langchain/pull/34458
Release Notes
https://github.com/langchain-ai/langchain/releases/tag/langchain-core%3D%3D0.3.81
Release Notes
https://github.com/langchain-ai/langchain/releases/tag/langchain-core%3D%3D1.2.5
More from langchain
9.8
CVE-2024-8309
SQL Injection in langchain-ai/langchain
8.3
CVE-2025-65106
LangChain Vulnerable to Template Injection via Attribute Access in Prompt Templates
8.2
CVE-2026-44843
LangChain: Unsafe deserialization of attacker-controlled LangChain objects through overly broad `load()` allowlists
7.5
CVE-2026-34070
LangChain Core has Path Traversal vulnerabilites in legacy `load_prompt` functions
7.5
CVE-2024-58340
LangChain <= 0.3.1 MRKLOutputParser ReDoS