CompTIA · Security+ SY0-801
AI threats and vulnerabilities
Security+ SY0-801 · Lesson 13 of 32 · 1:42:16
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Prepare for the Security+ SY0-801 exam (CompTIA Security+ V8). In this lesson: attacks on models (model manipulation, data poisoning, and evasion), prompt injection and jailbreaking, unreliable output, data loss through AI tools, and risky actions by AI agents, ending with Ashgrove's AI plan. Includes three original practice questions.
Preview the topics
- A buyer asks Ashgrove’s assistant to summarize a supplier’s price list. Hidden in white text on the document is a line that tells the assistant to email the buyer’s contacts to an outside address, and the assistant tries to do it. Which attack is this?
- A customer asks Ashgrove’s chatbot about the warranty on a toaster, and the chatbot confidently promises ten years. Ashgrove’s real warranty is two years, and the policy says nothing about ten. What is this?
- Ashgrove’s internal assistant can search every document in the company, and it shows a warehouse employee the salary list of the finance team. Which weakness caused this?