TL;DR
OWASP defined the Top 10 web security threats. CVSS defined how to score vulnerabilities. Lighthouse defined how to measure web performance.
Nobody defined: "What's your organization's overall AI security + visibility + data protection score?"
So we did.
AI Application Security Standard (AASS) v1.0 — three dimensions, one score:
AI Trust Score = AVS × 0.30 + PDS × 0.35 + ADP × 0.35
| Dimension | What it answers | Checks |
|---|---|---|
| AVS (AI Visibility) | Can AI find you? | 92 checks (SEO + AEO + AAO) |
| PDS (Prompt Defense) | Is your AI hardened? | 12 attack vectors |
| ADP (Data Protection) | Are you leaking PII to AI? | 15+ detection types |
All open source. All free. All deterministic.
Full spec: github.com/ppcvote/avs-standard/spec/AASS-v1.0.md
Free scanner: ultralab.tw/probe
"We don't sell security. We define it." — Ultra Lab
Originally published on Ultra Lab — we build AI products that run autonomously.
Try UltraProbe free — our AI security scanner checks your website for vulnerabilities in 30 seconds: ultralab.tw/probe
Top comments (0)