What is adversarial machine learning and why does it matter?
Answered
Adversarial machine learning exploits fundamental vulnerabilities in ML components — hardware, software, workflows and supply chains — to cause unintended behaviours such as degraded performance, unauthorised actions or extraction of sensitive model information. Examples include prompt injection and data poisoning. Treat these alongside standard cyber threats.
From the guidance
Primary (how) Guidelines for secure AI system development
As well as existing cyber security threats, AI systems are subject to new types of vulnerabilities. The term ‘adversarial machine learning’ (AML), is used to describe the exploitation of fundamental vulnerabilities in ML components, including hardware, software, workflows and supply chains.
Section: Why is AI security different?
Read this in Guidelines for secure AI system development (opens in new tab)