Loading lesson

AI and LLM threats and vulnerabilities (SY0-801)

Domain 2: threats, vulnerabilities, and mitigations (V7 22%) / attacks (V8 24%) · Lesson 8 of 8

Lesson 8 of 8

Ask a tutor

Lesson 8 of 8 · Domain 2: threats, vulnerabilities, and mitigations (V7 22%) / attacks (V8 24%)

AI and LLM threats and vulnerabilities (SY0-801)

3 quick checks≈2% of the testCore

Reading is free. Sign in to tick off lessons, keep your place and track your mastery.

By the end you’ll be able to

Explain model manipulation, training-data poisoning and evasion attacks against AI systems
Recognise direct and indirect prompt injection and jailbreaking against LLM applications, and the session hijacking or code execution they can lead to
Assess data loss and privacy risks when users or applications send sensitive data to AI tools

0% read

AI and LLM threats and vulnerabilities (SY0-801)Reading · 0%

Evaluate hallucinations, bias, explainability and ethical considerations as security and governance risks

Treat LLMs as an attack surface and select mitigations: input/output filtering, least-privilege tool access, human review, data classification and an AI acceptable-use policy

This lesson hasn’t been written yet

Security+ lessons are written from the official syllabus the first time a learner opens them, then saved for everyone. Sign in and this one is written for you in a minute or two. It’s free.

Sign in Create a free account