CVE-2026-24055

Affects: large language model

CVSS
MEDIUM · 5.3 v3.1
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N
Published
2026-01-22
Weakness
CWE-284, CWE-862
Source
nvd.nist.gov/vuln/detail/CVE-2026-24055

Description

Langfuse is an open source large language model engineering platform. In versions 3.146.0 and below, the /api/public/slack/install endpoint initiates Slack OAuth using a projectId provided by the client without authentication or authorization. The projectId is preserved throughout the OAuth flow, and the callback stores installations based on this untrusted metadata. This allows an attacker to bind their Slack workspace to any project and potentially receive changes to prompts stored in Langfuse Prompt Management. An attacker can replace existing Prompt Slack Automation integrations or pre-register a malicious one, though the latter requires an authenticated user to unknowingly configure it despite visible workspace and channel indicators in the UI. This issue has been fixed in version 3.147.0.

References

How GTK Cyber trains on this

AI security training at GTK Cyber covers the LLM and ML-pipeline vulnerability classes that vulnerabilities like CVE-2026-24055 fall into. Our hands-on courses are taught by Charles Givre and other practitioners who break and defend production AI systems.

AI Red-Teaming course → · Browse MITRE ATLAS techniques

Related AI/LLM CVEs

AI security training, taught by people who do the work.

Hands-on courses on adversarial AI, prompt injection, and ML pipeline security.

Explore AI Red-Teaming