Why randomness matters now

A random number generation weakness in the CryptoJS library meant that values intended to be unguessable could be reproduced. Because the library is embedded in a very large number of downstream projects, applications inherited the weakness without any mistake of their own.

Public reporting linked the flaw to wallet applications whose users lost funds — the practical outcome of a secret that could be recomputed rather than guessed.

A hardware wallet vendor shipped a hotfix after a code path that was meant to read the device's hardware random source resolved instead to a software generator. The device had good hardware entropy available; the software simply did not reach it.

This is the most common shape of real-world entropy failure: not an absent source, but an unverified path between the source and the code that consumes it.

Security researchers analysing wallet seed generation make a point that generalises well beyond cryptocurrency: patching a faulty generator protects future secrets only. Keys, seeds and certificates already created from the weak source remain predictable for as long as they exist.

Remediation therefore means finding, revoking and reissuing — which for a fleet of shipped devices may be impossible.

Standardised post-quantum algorithms are published, and national authorities have moved from encouragement to timetables for migrating away from classical public-key cryptography.

Those algorithms still generate their keys from random input. A migration programme is the natural moment to ask where that input comes from and whether its quality can be demonstrated under frameworks such as NIST SP 800-90B or BSI AIS 31.