The Ledger That Scores
Coverage. Not Claims.
LLM Threat scores 27 LLM-security vendors and 15 open-source tools against the OWASP LLM Top 10 (2026) and MITRE ATLAS. Each profile marks all 10 risks as Covered, Partial, Not covered, or Unverified, and every one of those cells was checked and dated by hand. Last verified 2026-10-30.
- 22/27
- Cover LLM01 Prompt Injection
- 1/27
- Cover LLM08 Hidden Context Exposure
- 8/27
- Vendors acquired
- 270
- Cells scored by hand
- 27,313★
- Stars on promptfoo, top tool
Every Vendor Site Sounds The Same. This Grid Shows What They Actually Cover.
Most vendor sites reprint marketing bullets. We map every vendor onto a fixed grid (ten OWASP risks, four plain verdicts) and name the exact feature behind each score.
"Not covered" and "Unverified" aren't the same verdict. Not covered means we checked and it isn't there. Unverified means the vendor claims it but no source confirms it — a question for the demo call, not a win.
| OWASP LLM Risk | Vendors covering it |
|---|---|
| LLM01 Prompt Injection | 22/27 |
| LLM02 Sensitive Information Disclosure | 20/27 |
| LLM03 Excessive Agency | 15/27 |
| LLM04 Supply Chain | 7/27 |
| LLM05 Data and Model Poisoning | 5/27 |
| LLM06 Unbounded Consumption | 5/27 |
| LLM07 Misinformation | 8/27 |
| LLM08 Hidden Context Exposure | 1/27 |
| LLM09 Vector and Embedding Weaknesses | 2/27 |
| LLM10 Improper Output Handling | 3/27 |
Of the ten risks, prompt injection has the best coverage — 22 of the 27 vendors we track cover it — while hidden context exposure is nearly untouched, with 1 covering it, per our review of public vendor documentation.
Six Ways Into One Grid
AI Red Teaming
Tools and services that attack your model before a real adversary does, hunting injection paths, jailbreaks, and data leaks.
5 vendors · 6 toolsGuardrails / LLM Firewall
Request-time filters that sit between users and models, blocking injected prompts, unsafe output, and PII leaks before they land.
8 vendors · 6 toolsAI-SPM / Runtime Protection
Posture management with live detection: who's running which model where, and whether anything is attacking it right now.
7 vendors · 0 toolsAI Governance / Model Risk
Inventory, policy mapping, evidence, and audit workflows for GRC and model-risk teams — governance rather than guardrails.
1 vendors · 0 toolsAgentic AI Security
Controls for AI that acts — tool-calling agents and MCP traffic — where excessive agency turns a bad prompt into a bad action.
6 vendors · 3 toolsOpen-Source Tools
Public GitHub projects for testing and defending LLMs, tracked with live stars, license, and repo status so a dead project can't sneak into your stack.
15 toolsSources: OWASP LLM Top 10 (2026), MITRE ATLAS, vendor documentation, GitHub API. See Data & Sources. Last verified 2026-10-30.