CVE-2023-38896 LangChain vulnerable to arbitrary code execution

Critical 9.8 Published 15 Aug 2023 Updated 9 Oct 2024

Remote attackers can execute arbitrary code via specific prompt functions in the LangChain library, requiring an upgrade to secure the application.

Exploited in the wild: Not known
Login needed: Not needed
Attackable: Over the internet
Victim action needed: Not needed
Fix: Available

What is CVE-2023-38896?

CVE-2023-38896 is a critical code injection vulnerability in n/a, published on 15 Aug 2023. An attacker can exploit it over the internet without logging in, with no action from the victim.

Official description

An issue in Harrison Chase langchain v.0.0.194 and before allows a remote attacker to execute arbitrary code via the from_math_prompt and from_colored_object_prompt functions.

Weakness: CWE-74

Which versions are affected by CVE-2023-38896?

3 open source packages ship an affected version, including langchain. It affects 1 product.

Package	Affected	Fixed in
`langchain` PyPI	All versions	`0.0.195`
`langchain` PyPI	All versions	`0.0.236`
`langchain` PyPI	All versions	`0.0.236`

Products include Langchain Langchain.

How do I fix CVE-2023-38896?

Upgrade langchain to a fixed version for your release line: 0.0.195 or 0.0.236.

Upgrade, and never evaluate templates, expressions or code built from user input. More about Code injection.

Is CVE-2023-38896 being exploited?

It is not confirmed as exploited in the wild, but public exploit code for CVE-2023-38896 exists.

Sources

Also known as GHSA-92j5-3459-qgp4, PYSEC-2023-146.