GreyNoise-observed LLM endpoint enumeration campaign
Campaign
Summary
Hide ▲
Show ▼
GreyNoise observed a December 28 campaign that generated 80,469 sessions over 11 days while probing more than 73 exposed or misconfigured LLM endpoints. The activity used OpenAI-compatible and Google Gemini API formats with harmless low-noise queries to avoid security alerts. The pattern indicates organized reconnaissance for commercial LLM access, though there is no confirmed post-discovery abuse in the evidence.
Related Happenings
Proton Meet launches privacy-focused encrypted conferencing service
Security Tool/Service
First: 01.04.2026 01:42
Last: 01.04.2026 01:42
Sources 1
About this happening:
**Proton Meet** launched as a **privacy-focused video conferencing service**, adding **end-to-end encrypted** calls for users who want an alternative to mainstream meeting platfor...
Proton Meet launches privacy-focused encrypted conferencing service
Security Tool/ServiceAbout this happening: **Proton Meet** launched as a **privacy-focused video conferencing service**, adding **end-to-end encrypted** calls for users who want an alternative to mainstream meeting platfor...
OpenAI Codex Security rolls out as a research-preview vulnerability-finding agent
Security Tool/Service
First: 07.03.2026 18:28
Last: 07.03.2026 18:28
Sources 1
About this happening:
**OpenAI** began rolling out **Codex Security** in **research preview**, adding an AI security agent that can **find, validate, and propose fixes** for vulnerabilities. The rollou...
OpenAI Codex Security rolls out as a research-preview vulnerability-finding agent
Security Tool/ServiceAbout this happening: **OpenAI** began rolling out **Codex Security** in **research preview**, adding an AI security agent that can **find, validate, and propose fixes** for vulnerabilities. The rollou...
Google API keys Gemini single-service privilege escalation privilege-escalation flaw
Vulnerability
First: 26.02.2026 22:55
Last: 26.02.2026 22:55
Sources 1
About this happening:
**Google API keys** exposed in public code became a **Gemini authentication weakness**, allowing copied keys to reach **private data** and incur **API charges** on victim accounts...
Google API keys Gemini single-service privilege escalation privilege-escalation flaw
VulnerabilityAbout this happening: **Google API keys** exposed in public code became a **Gemini authentication weakness**, allowing copied keys to reach **private data** and incur **API charges** on victim accounts...
Google Gemini leaked API key mitigation
Advisory/Mitigation
First: 26.02.2026 22:55
Last: 26.02.2026 22:55
Sources 1
About this happening:
**Google** is rolling out **mitigations for leaked API keys** that can reach **Gemini API** data, reducing the risk of unauthorized access and usage charges. New **AI Studio keys*...
Google Gemini leaked API key mitigation
Advisory/MitigationAbout this happening: **Google** is rolling out **mitigations for leaked API keys** that can reach **Gemini API** data, reducing the risk of unauthorized access and usage charges. New **AI Studio keys*...
DeepSeek, Moonshot AI, and MiniMax Claude capability-extraction campaign
Campaign
First: 24.02.2026 08:04
Last: 24.02.2026 08:04
Sources 1
About this happening:
**DeepSeek**, **Moonshot AI**, and **MiniMax** ran an **industrial-scale campaign** to extract **Claude** capabilities, using **fraudulent accounts** and **commercial proxy servic...
DeepSeek, Moonshot AI, and MiniMax Claude capability-extraction campaign
CampaignAbout this happening: **DeepSeek**, **Moonshot AI**, and **MiniMax** ran an **industrial-scale campaign** to extract **Claude** capabilities, using **fraudulent accounts** and **commercial proxy servic...
Timeline
-
09.01.2026 21:56 2 articles · 4mo ago
LLM endpoint enumeration begins on December 28
Campaign Scope UpdateOn December 28, 2025, a second campaign begins against exposed or misconfigured LLM endpoints, with two IP addresses systematically probing more than 73 model endpoints using OpenAI-compatible and Google Gemini API formats and low-noise queries intended to avoid security alerts.
Show sources
- Hackers target misconfigured proxies to access paid LLM services — www.bleepingcomputer.com — 09.01.2026 21:56
- Hackers target misconfigured proxies to access paid LLM services — www.bleepingcomputer.com — 09.01.2026 21:56
-
09.01.2026 21:56 1 articles · 4mo ago
GreyNoise discloses the LLM endpoint enumeration campaign
Initial DisclosureGreyNoise discloses that threat actors are hunting misconfigured proxy servers to reach paid commercial LLM services, describing an ongoing late-December campaign that probed more than 73 LLM endpoints and generated over 80,000 sessions using low-noise prompts to avoid security alerts.
Show sources
- Hackers target misconfigured proxies to access paid LLM services — www.bleepingcomputer.com — 09.01.2026 21:56