PHP2Shell : See if you’re affected with a free external attack surface scan. Click Here > 

Back
New CVE Detected

CVE-2025-66455 – Unauthenticated Remote Code Execution via Insecure Pickle Deserialization – LMDeplo

Be the first to know when new zero-days emerge:

Subscribe to Email Alerts
Subscribe to RSS Feed

Summary

CVE-2025-66455 is a critical remote code execution vulnerability in LMDeploy, an open-source LLM inference and serving toolkit maintained by InternLM. The flaw stems from unsafe pickle deserialization in the PyTorch DistServe control plane, where messages received over a ZeroMQ socket are deserialized without validation. An unauthenticated attacker who can reach the exposed /distserve/p2p_connect HTTP endpoint can trigger arbitrary code execution on the server, earning the flaw a CVSS v3.1 base score of 9.8 (Critical).

Technical details

  • Root cause: The EngineP2PConnection.handle_zmq_recv() handler (in lmdeploy/pytorch/disagg/conn/engine_conn.py) uses PyZMQ’s recv_pyobj(), which internally calls Python’s pickle.loads() on data received from a ZeroMQ PULL socket, without sanitizing or validating the payload before deserialization (CWE-502: Deserialization of Untrusted Data).
  • Trigger conditions: An attacker sends a request to the POST /distserve/p2p_connect endpoint, supplying an attacker-controlled remote_engine_endpoint_info.zmq_address. This causes the LMDeploy server to connect to the attacker’s ZeroMQ endpoint and deserialize a malicious pickle payload sent back, resulting in code execution.
  • Attack vector: Network-based, no authentication or user interaction required. The advisory notes that API-key authentication is not enabled by default and must be explicitly configured by the operator, meaning many deployments expose this endpoint without any access control.
  • Impact: Full compromise of confidentiality, integrity, and availability of the affected LMDeploy server, including arbitrary code execution in the context of the serving process — which may run with access to model weights, inference infrastructure, and any connected internal network resources.

Affected software

  • LMDeploy versions >= 0.9.2 and < 0.16.0 (i.e., 0.9.2 through 0.15.x)
  • Fixed in LMDeploy 0.16.0

Severity

  • CVSS v3.1 Base Score: 9.8 (Critical)
  • Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
  • CWE-502: Deserialization of Untrusted Data

Mitigation and recommended actions

  • Immediate: Upgrade LMDeploy to version 0.16.0 or later, which replaces the unsafe pickle-based send_pyobj()/recv_pyobj() calls with JSON serialization (send_json()/recv_json()) and Pydantic schema validation on the DistServe control plane.
  • If immediate patching is not possible:
    • Explicitly enable API-key authentication on LMDeploy serving endpoints, since it is not on by default.
    • Restrict network access to /distserve/* endpoints and any associated ZeroMQ control-plane ports to trusted hosts only (e.g., via firewall rules or network segmentation).
    • Do not expose LMDeploy DistServe endpoints directly to the public internet; place them behind a trusted internal network or VPN.
    • Block outbound connections from LMDeploy serving nodes to untrusted or attacker-influenceable ZeroMQ addresses where feasible.

References

Are you exposed?

Get a free report of your organization’s exposure to this CVE and threat

  • Mapping of all Assets with this Technology
  • Identification of potentially exposed assets to this CVE
  • Confirmation of verified exploitable assets
Get Exposure Report

How IONIX’s External Exposure Management Platform Detects and Validates
Zero-Days to Shrink MTTR

1

Map your entire attack surface (continously)

IONIX uses multi-factor discovery methods, including DNS analysis, certificate mapping, metadata inspection, and more, to automatically map every internet-facing asset across your environment. This includes cloud instances, third-party platforms, shadow IT, and even forgotten infrastructure that traditional tools miss.

2

Monitor for new CVEs

Dozens of threat intel feeds using agentic technology are continuously analyzed to detect the appearance of proof-of-concept code, exploit kits, and indicators of active targeting. IONIX goes further by applying AI to proactively evaluate whether emerging vulnerabilities are likely to be exploited, even before PoCs go public.

3

Identify Potential External Exposures

Not all CVEs matter. IONIX filters vulnerabilities by asking attacker-centric questions: Can it be reached from the internet? Does it require authentication? Is it being exploited in the wild? This dramatically reduces noise and focuses teams on threats that can actually be weaponized.

4

Create Safe, Scalable Exploit Validations

IONIX transforms real-world PoCs into safe, non-intrusive test payloads that can be run in production environments without disruption. These simulations are precisely targeted to the systems that are vulnerable, ensuring rapid validation without unnecessary load.

5

Execute Exploit Validations

By combining context about software stack, versioning, exposure status, and reachability, IONIX ensures that only the right payloads are executed against the right assets, maximizing efficiency and minimizing risk.

6

Drive Fast and Actionable Remediation

Results are routed through integrations with ticketing, SOAR, and SIEM tools. Issues are written in plain language, bundled into remediation clusters, and prioritized based on asset criticality, exploitability, and blast radius. This shortens mean time to remediation (MTTR) and empowers teams to act with confidence.

Are you exposed?

Get a free report of your organization’s exposure to this CVE and threat

  • Mapping of all Assets with this Technology
  • Identification of potentially exposed assets to this CVE
  • Confirmation of verified exploitable assets
Get Exposure Report
Close

Subscribe to Threat Center RSS

Copy/paste the link below into your preferred RSS reader or follow these instructions to subscribe to Slack alerts.

Open RSS feedSlack setup instructions

Close

Get Real-Time CVE Alerts to Your Email

Be the first to know when new zero-days emerge