CyberSafe AI¶
Abstract
This article provides an overview of the Cyber Range's CyberSafe AI platform, explaining its learning objectives, different play styles, available levels, and how to gain access to its materials.
What is CyberSafe AI?¶
CyberSafe AI is a freely available web-based platform that allows players to test and hone their artificial intelligence safety skills, covering the following learning objectives:
- Learn how to spot red flags in AI-generated content
- Learn how attackers use AI to craft convincing phishing messages and deepfakes
- Learn how to safely write prompts
- Learn how prompt injection and jailbreaking attacks work and what defenders do to stop them
- Learn how agentic AI systems can be manipulated, see what attackers might try when they exploit one, and how the right defenses keep things under control
CyberSafe AI offers two play styles and five different levels available to play through, each offering several multiple-choice and short-answer questions that provide players with real-time feedback and explanations.
How To Play¶
-
To access CyberSafe AI, players will need to navigate to the following link: CyberSafe AI

-
Next, players have the choice to authenticate with the Cyber Range using either Google or Azure AD, or sign in with a username/password.

-
Once logged in, players will be able to choose between two play styles: Story Mode or Level Select.

Story Mode¶
Story Mode allows players to play through all levels in order, providing a progressive story where players solve questions to earn enough stars to unlock the next level.

Level Select¶
Unlike Story Mode, where players earn points and progress through levels, Level Select allows players to jump to any level and practice the skills they need. All levels are available to play in any order.
Available Levels¶
With each level, the player's corgi buddy will gain more cyber swag.
-
Level 1 - Hallucination Hunter: Learn how to spot red flags in AI-generated content

-
Level 2 - Phish and Detect: Learn how attackers use AI to craft convincing phishing messages and deepfakes

-
Level 3 - Privacy Guardian: Learn what counts as personal information, why sharing too much can be risky, and how to rewrite prompts the safe way

-
Level 4 - Prompt Safety Expert: Learn how prompt injection and jailbreaking attacks work and what defenders do to stop them

-
Level 5 - Rogue Agent: Learn how agentic AI systems can be manipulated, see what attackers might try when they exploit one, and how the right defenses keep things under control

How Do I Gain Access to CyberSafe AI Materials?¶
Instructors can request CyberSafe AI supporting materials by filling out the following form:
Cyber Range CyberSafe AI Materials Request Form
Once you’ve been successfully verified by our Support Team, we will email you with further instructions!
Have a Question? Contact Support¶
Important
Students, please reach out to your instructor who can submit a ticket to our Support Team on your behalf.
We're here to help you. If you still have questions after reviewing the information above, please feel free to submit a ticket with our Support Team and we'll get back to you as soon as possible.
Thank You for Your Feedback!