Live Exposure Defense: From CVE to Confirmed Exposure in 12 Hours – See more

Back
New CVE Detected

CVE-2026-93839 – Authentication Bypass / Prompt Disclosure & SSRF – LightLLM ≤ 1.2.0

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

Subscribe to Email Alerts
Subscribe to RSS Feed

Summary

CVE-2026-93839 is a critical authentication bypass (CWE-306: Missing Authentication for Critical Function) affecting LightLLM, ModelTC’s Python-based LLM inference and serving framework, in all versions through 1.2.0. The flaw resides in the PD (Prefill-Decode) Master’s /pd_register WebSocket endpoint, which accepts node-registration requests without verifying the requester’s identity or validating that the claimed peer address actually belongs to the connecting client. The issue carries a CVSS v3.1 base score of 9.8 (Critical) and a CVSS v4.0 base score of 9.3 (Critical), reflecting its unauthenticated, network-exploitable nature and high impact on confidentiality, integrity, and availability.

Technical details

  • Root cause: The /pd_register WebSocket handler (register_and_keep_alive() in lightllm/server/api_http_pd.py) calls websocket.accept() and begins processing the client’s registration JSON before performing any authentication check. The PD Master trusts the client_ip_port value supplied in the JSON payload instead of validating it against the actual TCP peer address of the WebSocket connection.
  • Trigger conditions: Any unauthenticated party with network access to the PD Master’s WebSocket listener can open a connection to /pd_register and submit a crafted JSON registration message.
  • Attack vector: Network (remote, no authentication or user interaction required, low attack complexity).
  • Impact:
    • Prompt disclosure: An attacker can register a rogue prefill/decode node so that subsequent user prompts routed to the cluster are delivered in plaintext to the attacker-controlled socket.
    • Denial of service: Registering a node with the same address as a legitimate one silently overwrites and removes the legitimate node from the routing pool, disrupting inference traffic.
    • Server-side request forgery (SSRF): Attacker-supplied node addresses are interpolated directly into the PD Master’s health-check requests without validation, allowing it to be coerced into issuing HTTP requests to arbitrary internal network addresses.
  • Reported to be mechanistically distinct from the previously disclosed pickle-deserialization RCE on related PD-mode WebSocket endpoints (CVE-2026-26220); fixing that issue does not remediate this one.

Affected software

  • ModelTC LightLLM: all versions up to and including 1.2.0

Severity

  • CVSS v3.1: 9.8 (Critical) — CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
  • CVSS v4.0: 9.3 (Critical) — CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N

Mitigation and recommended actions

  • Immediate: As of this writing, no patched LightLLM release beyond 1.2.0 has been published that resolves this specific issue; the fix remains tracked as an open item in GitHub Issue #1576. Organizations should monitor the official ModelTC/LightLLM repository and releases page for a fix and upgrade as soon as one is available.
  • If no patch is available, apply the following network- and configuration-level mitigations:
    • Do not expose the PD Master’s /pd_register WebSocket endpoint to untrusted networks; restrict access to trusted internal hosts only via firewall/network segmentation.
    • Place the PD Master behind a service mesh or reverse proxy that enforces mutual TLS or a shared-secret authentication scheme before allowing node registration.
    • Monitor logs for unexpected or duplicate node registrations, especially registrations that replace previously active prefill/decode nodes.
    • Restrict or allowlist the destinations the PD Master’s health-check mechanism is permitted to contact, blocking requests to private/internal IP ranges (RFC 1918, link-local, loopback).
    • Treat any LightLLM deployment reachable from untrusted networks as high risk until an authenticated registration mechanism is confirmed in place.

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