Artifact GuideGLOBALFIPS 140-3

FIPS 140-3 Entropy and DRBG

A focused guide to the entropy-source and deterministic random bit generator evidence that can affect FIPS 140-3 module validation.

Use it to separate module design facts from CMVP caveat, Security Policy, ESV, and test-report evidence needs.

Author
Sorena AI
Published
May 9, 2026
Updated
May 9, 2026
Sections
5

Structured answer sets in this page tree.

Primary sources
4

Cited legal and guidance references.

Publication metadata
Sorena AI
Published May 9, 2026
Updated May 9, 2026
Overview

Entropy and DRBG review under FIPS 140-3 starts with how the cryptographic module obtains entropy, where the entropy source sits relative to the cryptographic boundary or TOEPP, and how the DRBG seed, internal state, and generated SSP strength are documented. This page does not treat an entropy source or DRBG as approved by default; it helps teams identify the evidence that must be checked before making a validation, procurement, or release claim.

Section 1

Classify how the module obtains entropy

Start by drawing the entropy path for each DRBG instantiation and reseed: source, interface, boundary, caller, and whether the module actively requests entropy or passively receives it. CMVP Implementation Guidance 9.3.A distinguishes modules that generate entropy themselves or call a well-defined GetEntropy() interface from modules that passively receive entropy through a seed loader, preloaded seed, I/O port, buffer, file, API, or other caller-supplied input.

That classification drives the evidence. For active, well-defined sources, the testing lab corroborates the vendor's entropy-strength estimate and the Security Policy states the minimum entropy generated or requested. For external or passive inputs, the Security Policy and certificate caveats may need to warn that the minimum strength of generated SSPs is not assured. If an entropy source is inside the TOEPP, the module cannot rely on code outside the cryptographic boundary to fetch and deliver entropy; the module needs direct access to the GetEntropy() interface.

  • Record whether each source is inside the cryptographic boundary, inside the TOEPP, outside the TOEPP, or externally loaded.
  • Document the exact interface used by the module: direct GetEntropy() call, callback with a direct path, GET command, LOAD command, seed loader, preloaded seed, or caller-supplied API buffer.
  • State the minimum number of entropy bits generated, requested, accepted, or believed to have been loaded for SSP generation.
  • Flag passive paths early because they can trigger certificate caveats or, for some inside-TOEPP designs without direct GetEntropy() access, prevent validation.
Section 2

Tie entropy claims to Security Policy and certificate caveats

The Security Policy should not merely say that a DRBG is seeded. It should state the entropy source scenario, the minimum entropy amount, the basis for the estimate, and the caveat text that applies when the lab cannot directly verify enough entropy for generated SSP strength. CMVP guidance also says the burden of proof is on the vendor when a tester reviews the rationale.

A practical review therefore compares three records side by side: the design description, the Security Policy language, and the module certificate caveat. If the design changed from active GetEntropy() access to externally loaded entropy, or if an operational environment moved the entropy source outside the tested perimeter, the certificate and Security Policy assumptions may no longer describe the deployed module.

  • For active entropy generation or a direct call to a known source, keep the vendor entropy-strength estimate and the lab corroboration evidence with the Security Policy text.
  • For externally requested or externally loaded entropy, confirm whether the certificate needs the no-assurance caveat and whether the Security Policy explains the caller responsibility.
  • For generated SSPs or random strings whose strength is limited by available entropy, make the caveat and Security Policy language match the affected output.
  • For configurable entropy sources, list the configured source, any ESV certificate number, and the Security Policy configuration rule that makes the source valid for the module.
Section 3

Protect SP 800-90A DRBG critical security parameters

For SP 800-90A DRBG mechanisms, CMVP guidance treats the entropy input string and seed as critical security parameters. It also identifies secret working-state values for Hash_DRBG, HMAC_DRBG, and CTR_DRBG. This matters for design reviews because a DRBG service can look like a simple random API while its seed and internal state still need CSP handling, boundary protection, and test-report coverage.

CTR_DRBG needs extra attention when the derivation function is not used. The test report must indicate whether a derivation function is used during instantiation and reseeding; without it, the report must demonstrate seeding by a full-entropy source, and CMVP guidance places that source inside the TOEPP or physical perimeter or cryptographic boundary for hardware modules.

  • List DRBG mechanism type: Hash_DRBG, HMAC_DRBG, CTR_DRBG, or another approved random bit generation construction in scope.
  • Classify entropy input, seed, and required internal-state values as CSPs in the Security Policy and design evidence.
  • Show how non-DRBG code and other DRBG instantiations are prevented from accessing DRBG internal state.
  • For CTR_DRBG, record whether the derivation function is used and attach the full-entropy source evidence when it is not.
Section 4

Check SP 800-90B and multiple-source evidence

When an entropy source must be assessed under SP 800-90B, the evidence should include more than a statistical-test output. CMVP guidance says the lab still performs heuristic analysis of the entropy source, and it points to the CMVP implementation of the SP 800-90B entropy assessment tool for statistical testing.

Multiple entropy sources should not be counted by intuition. CMVP guidance allows concatenating outputs from multiple independent SP 800-90B-compliant sources to provide a DRBG seed, but the Security Policy must explain the nature of the sources, which are creditable, and whether the certificate identifies all-physical or not-all-physical validated entropy sources. If sources are mutually dependent, at most one dependent source may be credited.

  • Keep raw entropy data handling, statistical-test results, and heuristic entropy-source analysis together in the validation evidence pack.
  • Distinguish physical and non-physical entropy sources before crediting entropy from multiple sources.
  • Credit only independent source behavior when summing entropy for a concatenated seed, and document any dependency argument.
  • Do not rely on external conditioning to rescue an entropy claim where current CMVP guidance does not allow it.
Section 5

Release checklist for entropy and DRBG claims

Use this checklist before relying on a FIPS 140-3 entropy or DRBG claim in a customer response, procurement review, certificate summary, or release gate. The goal is to make each public or internal claim traceable to module-specific evidence rather than a generic statement that the product uses an approved DRBG.

  • Module boundary: confirm the cryptographic boundary, TOEPP or physical perimeter, entropy source location, and DRBG service entry points match the validation evidence.
  • Entropy scenario: classify each source under the relevant CMVP entropy caveat scenario and record whether the module actively obtains entropy, passively receives it, or uses a hybrid design.
  • Security Policy: verify minimum entropy amounts, source configuration, caller guidance, caveat wording, random-string treatment, and ESV or ENT references.
  • DRBG CSPs: map entropy input, seed, and mechanism-specific internal state values to CSP handling, access controls, zeroization, and self-test evidence.
  • Algorithm evidence: attach CAVP or other validation records for the DRBG and related approved functions without implying that an algorithm certificate validates the whole module.
  • Change triggers: reopen the review after hardware, operating-system, virtualization, boundary, entropy-source, callback path, derivation-function, or DRBG-configuration changes.
Primary sources

References and citations

csrc.nist.gov
Referenced sections
  • CAVP records are useful evidence for approved algorithm implementations, but they do not replace module validation scope review.
github.com
Referenced sections
  • CMVP guidance cites this tool for SP 800-90B statistical testing while preserving the lab's responsibility for heuristic analysis.
Related guides

Explore more topics

FIPS 140-3 algorithm certificate mapping: ACVTS certificates to module boundary
Map CAVP algorithm certificates to FIPS 140-3 module services, approved security functions, security policy tables, and validation evidence.
FIPS 140-3 Algorithm Certificates FAQ
How CAVP algorithm certificates support, but do not replace, FIPS 140-3 cryptographic module validation evidence.
FIPS 140-3 Applicability Test
Check whether FIPS 140-3 applies to a cryptographic module claim by testing agency use, module boundary, security level, approved functions, CMVP status, and procurement evidence.
FIPS 140-3 Approved and Non-Approved Mode Workflow
Classify FIPS 140-3 module services by approved security service, allowed no-security-claimed use, and non-approved service evidence.
FIPS 140-3 approved-mode evidence workflow
A grounded workflow for collecting FIPS 140-3 approved-mode evidence: module boundary, approved services, service indicators, CAVP certificates, Security Policy entries, and change review.
FIPS 140-3 Certificate Maintenance FAQ
How to maintain FIPS 140-3 certificate evidence after validation by checking module status, version, caveats, Security Policy, and revalidation records.
FIPS 140-3 Change Impact Review
Review FIPS 140-3 module changes against boundary, version, operational environment, embedded module, software loading, CVE, and certificate evidence.
FIPS 140-3 compliance guide
A grounded FIPS 140-3 compliance guide for cryptographic module scope, security-level claims, CMVP validation evidence, and procurement review.
FIPS 140-3 Entropy Evidence FAQ
How FIPS 140-3 entropy evidence should document entropy source location, GetEntropy access, SP 800-90B testing, Security Policy text, and certificate caveats.
FIPS 140-3 FAQ for Cryptographic Modules
Answers to common FIPS 140-3 questions about scope, CMVP validation, algorithm certificates, module boundaries, approved mode, and validation evidence.
FIPS 140-3 Module Boundaries FAQ
Understand how FIPS 140-3 module boundaries affect cryptographic module scope, interfaces, software and firmware components, and bound or embedded validated modules.
FIPS 140-3 Module Boundary Selector Workflow
A FIPS 140-3 workflow for selecting a cryptographic module boundary, separating embedded and bound modules, and collecting CMVP validation evidence.
FIPS 140-3 operational environments FAQ
Learn what a FIPS 140-3 operational environment means for software, firmware, and hybrid cryptographic modules, and what evidence to check before relying on a validation claim.
FIPS 140-3 security levels: how to choose and evidence them
A practical FAQ on FIPS 140-3 security levels, module scope, CMVP evidence, bound or embedded modules, and common claim mistakes.
FIPS 140-3 Security Policy Template
Build a FIPS 140-3 module Security Policy with sections for boundary, roles, services, approved algorithms, SSP handling, self-tests, and CMVP evidence.
FIPS 140-3 Validation Checklist
Checklist for preparing a cryptographic module for FIPS 140-3 validation: boundary, levels, services, approved algorithms, entropy, tests, security policy, and change evidence.
FIPS 140-3 Validation Maintenance
Maintain FIPS 140-3 validation claims by checking module identity, certificate status, boundary changes, operational environments, and CAVP evidence.
FIPS 140-3 Validation Maintenance Change Workflow
A FIPS 140-3 workflow for triaging module changes against CMVP validation scope, Security Policy evidence, CAVP certificates, software loading, and CVE records.
FIPS 140-3 Vendor Affirmation FAQ
When vendor affirmation can support a FIPS 140-3 module claim, what it does not supersede, and which Security Policy, CAVP, CSTL, and test-report evidence to keep.
FIPS 140-3 vs ISO/IEC 19790 and ISO/IEC 24759
Compare FIPS 140-3 with ISO/IEC 19790 and ISO/IEC 24759 for cryptographic module validation scope, evidence, testing, and procurement claims.
FIPS 140-3: CMVP Lifecycle Timeline
Practical FIPS 140-3 guidance for CMVP Lifecycle Timeline: scope, controls, evidence, source-linked decisions, and implementation checkpoints.
FIPS 140-3: FIPS 140-2 vs FIPS 140-3
Compare FIPS 140-2 legacy references with FIPS 140-3 requirements, ISO/IEC 19790 alignment, CMVP testing evidence, and guidance mappings.
FIPS 140-3: Module Boundary and Service Mapping
Map a FIPS 140-3 cryptographic module boundary to services, approved algorithms, operational environments, and CMVP validation evidence.
FIPS 140-3: Module Boundary Selector
Select and document a FIPS 140-3 cryptographic module boundary across hardware, software, firmware, operational environment, services, and validation evidence.
FIPS 140-3: Operational Environment
FIPS 140-3 operational environment guidance for software, firmware, hybrid, CAVP certificate, EVM, and PAA/PAI validation claims.
FIPS 140-3: Security Levels Explained
Explain FIPS 140-3 Security Levels 1 through 4, what they cover, and how to document level claims for cryptographic module validation.
FIPS 140-3: step-by-step workflow for mapping algorithm certificates to CMVP modules
Map CAVP algorithm certificates to a FIPS 140-3 module by matching implementation identity, operational environment, module services, and security policy evidence.
How should teams handle approved mode under FIPS 140-3?
Answer the FIPS 140-3 approved-mode question with service-level indicators, Security Policy evidence, and limits on non-approved functions.