Generating Secure Random IVs and Nonces
What you'll learn: You'll discover how to properly generate cryptographically secure random values for IVs and nonces using the right tools.
The Core Problem
Remember from earlier lessons that IVs and nonces must be unpredictable to prevent attacks like nonce reuse in CTR/GCM modes. But where do these random values actually come from? Not all "random" is created equal—using weak randomness is like locking your door but leaving the key under the mat.
CSPRNGs: Your Secure Random Source
A Cryptographically Secure Pseudo-Random Number Generator (CSPRNG) is a special algorithm designed to produce values that are:
- Unpredictable: An attacker can't guess the next value even knowing previous ones
- Uniformly distributed: All possible values are equally likely
- Non-repeating: Extremely low probability of generating the same value twice
Think of a CSPRNG like a casino-grade dice roller versus shaking a pair of dice in your hand—one is engineered for true randomness, the other might have subtle patterns.
Platform APIs: The Right Way
Modern platforms provide battle-tested CSPRNGs:
- Python:
secrets.token_bytes()oros.urandom() - Node.js:
crypto.randomBytes() - Java:
SecureRandom - C#:
RandomNumberGenerator.GetBytes()
Never use general-purpose random functions like Math.random(), rand(), or random.random()—these aren't cryptographically secure.
Practical Generation
For a 128-bit (16-byte) IV in CBC mode, you'd request 16 bytes from your platform's CSPRNG. Generate a fresh, random value for every encryption operation. Store or transmit it alongside your ciphertext (IVs aren't secret, just unpredictable).
Key Takeaway: Always use your platform's CSPRNG APIs (like secrets.token_bytes()) to generate IVs and nonces—never use standard random functions, as they lack cryptographic security properties.