Entropy assurance
The case for measuring an entropy source while it runs rather than validating it once and assuming it holds, and what that measurement establishes.
What it covers
- What a physical model buys you that output statistics cannot
- Min-entropy estimation against a model, and the boundaries of the result
- The runtime loop: physical process, measurement, entropy estimate, health state, audit output
- Where continuous visibility into source health sits among the other inputs to a security posture
Request the material on entropy assurance
Tell us where to send it and a little about what you are building.
An engineer replies, not an autoresponder. Say what you are building and the answer covers your case as well as the general one.
← All subjectsWhy this is being asked now
The entropy source used to be an implementation detail. Four changes in the standards and regulatory record moved it into the certification path.
FIPS 203, 204 and 205 finalised
A key-encapsulation mechanism and two signature schemes, standardised. Post-quantum moved from selection into deployment.
SP 800-90C became final
The construction standard connecting a validated entropy source to an approved deterministic mechanism — making the relationship between the two explicit.
BSI updated AIS 20/31
Physical, hybrid and deterministic RNG evaluation stayed central to Common Criteria work, with new transition guidance.
ANSSI expects PQC in qualification
French guidance states post-quantum considerations should enter product qualification from 2027. It is a strong procurement signal, not a legal ban.
Ask for the material.
Tell us who has to be convinced — an assessor, an architecture board, a programme team — and the answer is written for them.