Researchers have introduced the Eticas AI Risk Taxonomy, a new framework designed to operationalize AI audits. Unlike previous taxonomies that merely catalog risks, this system bridges the gap between identifying a risk and executing a measurable test against an AI system. The framework was demonstrated on GPT-4-0314, measuring PII leakage under increasing adversarial conditions and assigning a systemic grade. The Eticas AI Risk Taxonomy v2.0.0 organizes 76 subcategories across 10 categories and is published as open semantic infrastructure under CC BY 4.0. AI
IMPACT Provides a standardized, operational framework for AI auditing, enabling more consistent and measurable risk assessments across different systems.
RANK_REASON This is a research paper detailing a new taxonomy and methodology for AI auditing. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
