What Adversarial Machine Learning: A Taxonomy and Terminology of Attacks and Mitigations requires
This NIST Trustworthy and Responsible AI report develops a taxonomy of concepts and defines terminology in the field of adversarial machine learning (AML). The taxonomy is built on surveying the AML literature and is arranged in a conceptual hierarchy that includes key types of ML methods and lifecycle stages of attack, attacker goals and objectives, and attacker capabilities and knowledge of the learning process. The report provides corresponding methods for mitigating and managing the consequences of attacks, meant to inform standards and practice guides for assessing and managing AI system security by establishing a common language for the AML landscape. The data-driven approach of machine learning introduces security and privacy challenges beyond classical threats. These include the potential for adversarial manipulation of training data, adversarial exploitation of model vulnerabilities, and malicious interaction with models to exfiltrate sensitive information. AML is concerned with studying the capabilities of attackers and their goals, the design of attack methods that exploit ML vulnerabilities during the development, training, and deployment phases, and the design of ML algorithms that can withstand these challenges. The taxonomy of AML is defined with respect to five dimensions of risk assessment: AI system type, stage of the ML lifecycle process, attacker goals, attacker capabilities, and attacker knowledge.
Pillar: AI Governance & Law · Authority: National Institute of Standards and Technology · Version: 1.0.0 · Last updated:
Primary source: https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.100-2e2023.pdf
SHA-256 integrity: 9a6fca3511a16c4d8702155e7fcf50bb8f731dc93e54173c92c2ade551afdbfc
Primary Citations — 8 traced to source
- {"citation":"Executive Summary: AML is concerned with studying the capabilities of attackers and their goals, as well as the design of attack methods that exploit the vulnerabilities of ML during the development, training, and deployment phase of the ML lifecycle.","doc_section":"Executive Summary, Page 1"}
- {"citation":"Section 2.1.1: Adversarial machine learning literature predominantly considers adversarial attacks against AI systems that could occur at either the training stage or the ML deployment stage.","doc_section":"2.1.1. Stages of Learning"}
+ 6 more citations (full bibliography, deterministic workflow, actionable schema and crosswalks) included in the vault unlock — $0.01 via Skyfire / L402 / Direct Base USDC.
Access
- Discovery (free): /api/v1/nodes/nist-ai-100-2-aml-taxonomy.json — 6-field metadata
- Vault (full node): /api/v1/vault/nodes/nist-ai-100-2-aml-taxonomy.json — full 13-key payload, $0.01 USDC (L402/Skyfire/Direct Base)
- Canonical URL: https://bidda.com/intelligence/nist-ai-100-2-aml-taxonomy
- Back to registry: Browse all 10,108 compliance nodes