CVE-2026-103270

HIGHCVSS 7.5/10

Last modified

CVE-2026-103270 is a high-severity vulnerability rated 7.5/10 on the CVSS scale. LightLLM through 1.2.0 mounts reinforcement learning control routes on the public HTTP API without authentication checks. Unauthenticated attackers can call endpoints like /pause_generation, /abort_request, /flush_cache, and /init_weights_update_group to disrupt inference operations and wedge workers on deployments started with --enable_rl..

Description

LightLLM through 1.2.0 mounts reinforcement learning control routes on the public HTTP API without authentication checks. Unauthenticated attackers can call endpoints like /pause_generation, /abort_request, /flush_cache, and /init_weights_update_group to disrupt inference operations and wedge workers on deployments started with --enable_rl.

Metrics

Weakness Enumeration

References

Timeline

Published
Last Modified
Status
Deferred

Frequently Asked Questions

What is CVE-2026-103270?
LightLLM through 1.2.0 mounts reinforcement learning control routes on the public HTTP API without authentication checks. Unauthenticated attackers can call endpoints like /pause_generation, /abort_request, /flush_cache, and /init_weights_update_group to disrupt inference operations and wedge workers on deployments started with --enable_rl.
How severe is CVE-2026-103270?
CVE-2026-103270 has a CVSS score of 7.5/10 (HIGH severity).
How do I fix CVE-2026-103270?
Check the vendor references and advisories linked above for patched versions and mitigation guidance. You can also run a Strix scan to test if your systems are affected.

How Strix Helps

Related CVEs from 2026

Are you affected by CVE-2026-103270?

Run a free Strix scan to check your systems for this vulnerability.

Scan your code now

Source: NVD / NIST