CVE-2025-68664: LangChain Serialization Injection Vulnerability Allows Secret Extraction

Background

LangChain is an open-source framework designed for building applications that utilize language models. It provides functionalities for serialization and deserialization of objects, enabling developers to manage complex data structures easily. However, a critical vulnerability has been identified in its serialization methods, specifically in the dumps() and dumpd() functions, which fail to properly escape user-controlled data containing specific keys, leading to potential security risks.

Vulnerability Detail

A serialization injection vulnerability exists in LangChain's dumps() and dumpd() functions. These functions do not escape dictionaries with 'lc' keys when serializing free-form dictionaries. As a result, user-controlled data containing this key structure is treated as legitimate LangChain objects during deserialization. This vulnerability allows attackers to inject malicious structures through user data, enabling them to extract sensitive information such as environment variable secrets or instantiate classes with controlled parameters, potentially triggering harmful side effects.

pip install --upgrade "langchain-core==1.7.0"

Workaround

If immediate patching is not feasible, consider implementing input validation to sanitize user-controlled data before serialization. For example, avoid using user data directly in serialization functions or implement a custom serialization method that escapes or filters out 'lc' keys. Here is a simple code snippet to demonstrate input validation:

def safe_dumps(data):
    if isinstance(data, dict) and 'lc' in data:
        raise ValueError('Invalid data structure: contains forbidden key')
    return dumps(data)

Use safe_dumps() instead of dumps() to prevent serialization of unsafe data.

Conclusion

The LangChain serialization injection vulnerability (CVE-2025-68664) poses significant risks to applications using the affected serialization methods. It is crucial for developers to update to the latest patched version of the package and implement necessary security measures to protect against potential exploits.

See every vulnerability recorded for langchain-core →

FAQ