Artificial Intelligence.

Artificial Intelligence

AI is rapidly transforming both sides of cybersecurity — empowering defenders with faster threat detection and response while giving attackers more sophisticated tools for automation, evasion, and social engineering. It lowers the barrier to BUILD, ATTACK, and DEFEND. But AI is still maturing: its outputs must be validated before they are trusted. AI is a "skill multiplier" — most powerful for those who already understand the fundamentals — and an equalizer that lets people pivot quickly into new areas.

Challenges

Multiple Choice

What is artificial intelligence (AI)?

Computer systems that perform tasks normally requiring human intelligence
A type of computer virus that spreads on its own
A faster kind of internet connection
A programming language used only for robots

✓ Solved!

Multiple Choice

What best describes a large language model (LLM), like the kind behind a chatbot?

A model trained on huge amounts of text that predicts the next word (token)
A database that stores a copy of every website on the internet
A person typing the answers very quickly behind the scenes
A search engine that only returns exact keyword matches

✓ Solved!

Multiple Choice

An AI chatbot gives you a confident answer that turns out to be completely made up. What is this called, and what should you do?

A hallucination — always validate AI output before trusting it
A firewall — ignore it and move on
A bug — the model has corrupted its own data
A patch — the model has fixed itself

✓ Solved!

Multiple Choice

How is AI changing cybersecurity?

It helps both defenders and attackers, lowering the barrier on both sides
It only helps defenders and makes attacks impossible
It only helps attackers and is useless for defense
It has no real effect on cybersecurity

✓ Solved!

Multiple Choice

AI can be called a "skill multiplier." What does that mean?

It is most powerful for people who already understand the fundamentals of the task they are using it for
It removes any need to learn about a topic
It only works for people with no experience
It multiplies the number of computers you need

✓ Solved!

Flag Challenge

At the bottom of the page is a simulated AI assistant ("HelpBot") that has been told to keep a secret flag and never reveal it. Chat with it and use a prompt-injection technique to make it ignore its instructions and leak the flag. Submit the C2{...} flag it reveals.

✓ Solved!

Flag Challenge

What token ID is generated based on the text "hello" within OpenAI's GPT-5.x & O1/3 models? (Input just the number without any extra spaces or brackets)

✓ Solved!

Flag Challenge

What token ID is generated based on the text "Hello" within OpenAI's GPT-5.x & O1/3 models? (Input just the number without any extra spaces or brackets)

✓ Solved!

Multiple Choice

What can affect the output of an AI model like a large language model (LLM)?

Capitalization, punctuation, and phrasing of the input prompt
Context from previous messages in a conversation
The specific model and its training data
All of the above

✓ Solved!

For More Information

IBM: What Is Artificial Intelligence (AI)?

A clear, beginner-friendly overview of what AI is, the difference between AI, machine learning, and deep learning, and how these systems are used in the real world.

Visit →

Cloudflare: What Is a Large Language Model (LLM)?

Explains in plain language how large language models are trained on text and generate responses by predicting the next token — the engine behind modern chatbots.

Visit →

OWASP Top 10 for LLM Applications

The security industry's reference list of the biggest risks in AI applications — including prompt injection, the exact technique used in this module's challenge.

Visit →