CVE-2025-67729 lmdeploy vulnerable to Arbitrary Code Execution via Insecure Deserialization in torch.load()

What is CVE-2025-67729?

CVE-2025-67729 is a high-severity unsafe deserialization vulnerability in InternLM lmdeploy, published on 26 Dec 2025. An attacker can exploit it over the internet without logging in, and only if a user opens a file or follows a link.

Official description

LMDeploy is a toolkit for compressing, deploying, and serving LLMs. Prior to version 0.11.1, an insecure deserialization vulnerability exists in lmdeploy where torch.load() is called without the weights_only=True parameter when loading model checkpoint files. This allows an attacker to execute arbitrary code on the victim's machine when they load a malicious .bin or .pt model file. This issue has been patched in version 0.11.1.

Weakness: CWE-502

Which versions are affected by CVE-2025-67729?

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

Products include Internlm Lmdeploy.

How do I fix CVE-2025-67729?

Upgrade lmdeploy to 0.11.1 or later.

Upgrade, and only deserialise data from trusted sources or with an allow-list of types. More about Unsafe deserialization.

Is CVE-2025-67729 being exploited?

It is not confirmed as exploited in the wild, but public exploit code for CVE-2025-67729 exists.

Other vulnerabilities in InternLM lmdeploy

• CVE-2026-63764 LMDeploy Server-Side Request Forgery via HTTP Redirect Bypass High 7.7
• CVE-2026-33626 LMDeploy Vulnerable to Server-Side Request Forgery (SSRF) via Vision-Language Image Loading High 7.5
• CVE-2025-3162 InternLM LMDeploy PT File utils.py load_weight_ckpt deserialization Medium 4.8
• CVE-2025-3163 InternLM LMDeploy conf.py open code injection Medium 4.8

Sources

• CVE record
• NVD
• OSV
• GitHub advisory
• How this page is made

Also known as GHSA-9pf3-7rrr-x5jh, PYSEC-2026-1578.