Security categorisation is the first decision in the federal risk process and the one most often rushed, yet it sets which SP 800-53 baseline applies and therefore the cost and depth of every control that follows. FIPS 199 defines how to rate the potential impact of a loss of confidentiality, integrity and availability as low, moderate or high; FIPS 200 turns that rating into a system impact level using the high-water mark and points to the SP 800-53 baselines; and privacy runs in parallel through a separate privacy control baseline and the PII confidentiality impact level from SP 800-122. This guide explains each step with the official notation and examples, shows where privacy controls in the PT family fit, covers how identity systems and credentials should be categorised, and sets out a repeatable process with evidence. It is for system owners, ISSOs, privacy officers and assessors working with US federal systems, FedRAMP offerings or contracts that adopt NIST categorisation. For baselines and tailoring after categorisation see the NIST SP 800-53B Control Baselines and Tailoring Guide.
Key takeaways
- Categorise information types first, then the system. FIPS 199 rates each information type for confidentiality, integrity and availability; the system takes the highest value for each objective across all its information types.
- One high makes the system high. Under FIPS 200 a system is low only if all three objectives are low, moderate if at least one is moderate and none is high, and high if any one is high.
- Privacy has its own baseline. SP 800-53B defines a single privacy control baseline that applies irrespective of impact level, alongside the low, moderate and high security baselines.
- PII adjusts confidentiality, it does not replace it. SP 800-122 describes a PII confidentiality impact level, distinct from FIPS 199, that supplements the provisional confidentiality level using six factors.
- Identity infrastructure inherits from what it protects. An identity provider, secrets vault or CI/CD credential store that can reach a high system should be categorised with that in mind, because its integrity loss becomes every dependent system's loss.
Where categorisation sits
| Document | Status as published | Role in categorisation |
|---|---|---|
| FIPS 199, Standards for Security Categorization of Federal Information and Information Systems | Final, February 2004; CSRC shows no revision | Defines impact levels and the security category notation |
| FIPS 200, Minimum Security Requirements for Federal Information and Information Systems | Final, March 2006 | Sets 17 minimum requirement areas, the high-water mark and the link to SP 800-53 baselines |
| SP 800-60 Vol. 1 Rev. 1 and Vol. 2 Rev. 1 | Final, August 2008 | Catalogue of information types with recommended provisional impact levels |
| SP 800-53 Rev. 5 (RA-2 Security Categorization) | Final, with later releases | Control requiring categorisation to be done, documented and approved |
| SP 800-53B | Final, December 2020; release 5.2.0 (27 August 2025) made no baseline changes | Low, moderate and high security baselines and the privacy baseline |
| SP 800-122 | Final, April 2010 | PII confidentiality impact level and its factors |
| SP 800-37 Rev. 2 | Final, December 2018 | Places categorisation in the RMF Categorize step |
Scope: FIPS 199 and FIPS 200 apply to all federal information except classified information and to all federal information systems except national security systems; FIPS 200 encourages state, local and tribal governments and private sector critical infrastructure to use them. National security systems use CNSSI 1253, as the RA-2 guidance in SP 800-53 notes. For where this sits in the wider RMF, see the NIST Risk Management Framework (SP 800-37) Guide.
FIPS 199: impact levels and the security category
FIPS 199 defines three potential impact levels for each security objective. Low means the loss could be expected to have a limited adverse effect on organisational operations, organisational assets or individuals; moderate, a serious adverse effect; high, a severe or catastrophic adverse effect. The objectives are confidentiality (preserving authorised restrictions on access and disclosure, including privacy and proprietary information), integrity (guarding against improper modification or destruction, including non-repudiation and authenticity) and availability (timely and reliable access to and use of information).
The generalised notation is: SC information type = {(confidentiality, impact), (integrity, impact), (availability, impact)}. Acceptable values are LOW, MODERATE, HIGH or NOT APPLICABLE, and NOT APPLICABLE may be used only for confidentiality. A system can never be rated NOT APPLICABLE.
| FIPS 199 example | Confidentiality | Integrity | Availability |
|---|---|---|---|
| Public information on a web server | NA | MODERATE | MODERATE |
| Investigative information | HIGH | MODERATE | MODERATE |
| Routine administrative information | LOW | LOW | LOW |
| Contracting acquisition system | MODERATE | MODERATE | LOW |
| SCADA system (final category in the example) | MODERATE | HIGH | HIGH |
The public web server example is worth remembering: public content needs no confidentiality protection, yet defacement (integrity) and outage (availability) still carry moderate impact.
Information types and SP 800-60
SP 800-60 Vol. 1 Rev. 1 provides the guidelines for mapping types of information to security categories, and its appendices give categorisation recommendations and rationale for mission-based and management and support information types. Treat those recommendations as provisional: the system owner adjusts them for the specific system, documents the reason and has the adjustment approved.
- Identify all information types: include information the system stores, processes and transmits, and system information such as audit logs, configuration data and credentials, which are often forgotten.
- Start from provisional levels: select the SP 800-60 type that best matches, or define an organisational type where none fits.
- Adjust with context: aggregation (many records together), criticality to a mission, legal obligations and special handling can raise a level; reducing below the provisional level needs a clear rationale.
- Record the rationale: assessors will ask why each level was chosen, particularly where a level was lowered.
FIPS 200: from system category to baseline
FIPS 200 requires the impact determination before control selection. The system's overall impact level is the highest of the three objective values (the high-water mark), and organisations then select the matching SP 800-53 baseline, meet its assurance requirements and tailor it with approval documented in the system security plan. FIPS 200 states that there is no provision under FISMA for waivers.
| Example system category | High-water mark | Security baseline |
|---|---|---|
| {(C, LOW), (I, LOW), (A, LOW)} | Low | SP 800-53B low baseline |
| {(C, MODERATE), (I, LOW), (A, LOW)} | Moderate | Moderate baseline |
| {(C, LOW), (I, HIGH), (A, MODERATE)} | High | High baseline |
| {(C, NA), (I, MODERATE), (A, MODERATE)} information type on its own | System still needs C rated at least LOW | Moderate baseline |
The 17 FIPS 200 minimum requirement areas match SP 800-53 families: Access Control; Awareness and Training; Audit and Accountability; Certification, Accreditation, and Security Assessments; Configuration Management; Contingency Planning; Identification and Authentication; Incident Response; Maintenance; Media Protection; Physical and Environmental Protection; Planning; Personnel Security; Risk Assessment; System and Services Acquisition; System and Communications Protection; and System and Information Integrity. FIPS 200 predates the privacy and supply chain families added in SP 800-53 Rev. 5, which is one reason the privacy baseline sits outside the impact-level model.
Privacy controls: the privacy baseline and the PT family
SP 800-53 Rev. 5 integrated privacy controls into the main catalogue, and SP 800-53B publishes a privacy control baseline that applies irrespective of the system's security impact level. In practice organisations apply it to systems that process personally identifiable information (PII), as determined with the senior agency official for privacy. The PT family, Personally Identifiable Information Processing and Transparency, carries the privacy-specific requirements.
| Control | Title | What it adds to categorisation work |
|---|---|---|
| PT-2 | Authority to Process Personally Identifiable Information | Confirms the legal authority for each PII processing activity identified while listing information types |
| PT-3 | Personally Identifiable Information Processing Purposes | Documents why each PII type is processed, across the full life cycle including logging and analysis |
| PT-4 | Consent | Where consent is the basis for processing |
| PT-5 | Privacy Notice | Notice to individuals about processing |
| PT-6 | System of Records Notice | US Privacy Act systems of records |
| PT-7 | Specific Categories of Personally Identifiable Information | Extra protections for categories that law or risk assessment marks as particularly sensitive |
| PT-8 | Computer Matching Requirements | Matching programmes under US law |
| RA-8 | Privacy Impact Assessments | The analysis and document that show privacy was considered from the start |
| SI-12(1) | Limit Personally Identifiable Information Elements | Minimisation, which can lower the PII confidentiality impact |
PII confidentiality impact level (SP 800-122)
SP 800-122 describes a PII confidentiality impact level that is distinct from the FIPS 199 confidentiality level and is used to supplement the provisional confidentiality level. Its six factors are identifiability, quantity of PII, data field sensitivity, context of use, obligation to protect confidentiality, and access to and location of PII. They are weighed together, and the guide notes that a small quantity of PII should not by itself justify a lower impact level. Where PII needs no confidentiality protection, such as a public staff phone directory, the PII level is not applicable.
For a programme view of privacy alongside security, see the NIST Privacy Framework 1.1 Guide.
Where identity and NHI fit
Categorisation usually focuses on business data. Identity data and machine credentials deserve their own lines in the information type list, because their compromise changes the impact on everything they unlock. The points below are NHI Mgmt Group's practitioner guidance, not text from FIPS 199 or SP 800-60.
- Credentials and keys as information types: list service account passwords, API keys, signing keys, certificates and tokens. Their integrity and confidentiality impact should reflect the highest system they grant access to, not the system that stores them.
- Shared identity services: an identity provider, PAM platform or secrets vault serving a high system should be assessed for a high integrity rating; a forged token or altered policy is a compromise of the dependent system. See the Secrets Management Guide.
- Pipelines and automation: CI/CD systems with deployment rights to a high system carry that system's integrity impact. See the CI/CD Pipeline Identity Security Guide.
- Identity data as PII: directory attributes, authentication logs and biometric templates are PII; include them in PT-3 purposes and the SP 800-122 assessment. See the Identity Data Privacy and Consent Guide.
- Controls that follow: IA-4 Identifier Management, IA-5 Authenticator Management and AC-6 Least Privilege are all in the baselines; categorisation decides which enhancements apply.
Running a categorisation, step by step
- Define the boundary. Agree what is inside the system, including shared services it depends on and interconnections.
- List information types. Business data, system data (logs, configuration), identity data and credentials. Map each to an SP 800-60 type or an organisational type.
- Assign provisional levels. Use SP 800-60 recommendations, then adjust for aggregation, mission criticality and legal obligations.
- Assess PII. With the privacy officer, identify PII, apply the SP 800-122 factors and supplement the confidentiality level; decide whether the privacy baseline applies and whether a privacy impact assessment (RA-8) is needed.
- Apply the high-water mark. Take the highest value per objective for the system category and the overall impact level for baseline selection.
- Review and approve. The authorising official or designated representative reviews and approves the category, as RA-2 requires; record it in the system security and privacy plan (PL-2).
- Select and tailor baselines. Apply the security baseline for the impact level and the privacy baseline where applicable.
- Revisit on change. New information types, new interconnections, aggregation of data or a change of mission all trigger a re-categorisation.
Decision guidance: over-categorising a system to high adds significant control cost and slows delivery, while under-categorising is a finding that can invalidate an authorisation. If a single information type drives the system to high, consider whether it can be separated into its own boundary. RA-2(1) Impact-level Prioritization lets organisations sub-categorise moderate systems for finer prioritisation where that helps. Small agencies with a few systems may categorise centrally; large ones benefit from an organisational information type register reused across systems.
Evidence and measurement
- Categorisation currency: percentage of systems whose category was reviewed within the policy period or after the last significant change.
- Rationale completeness: share of information types with a documented rationale, especially where provisional levels were lowered.
- PII coverage: percentage of PII-processing systems with a PII confidentiality assessment and, where required, a privacy impact assessment.
- Credential types captured: share of systems whose information type list includes credentials and identity data.
- Re-categorisation findings: number of assessments where the assessor disagreed with the category; aim for zero.
Framework mapping
| Activity | NIST SP 800-53 Rev. 5 | NIST CSF 2.0 |
|---|---|---|
| Security categorisation | RA-2 Security Categorization; RA-2(1) Impact-level Prioritization | ID.AM-05 assets prioritised by classification, criticality and impact |
| Data inventory | CM-8 System Component Inventory (system side) | ID.AM-07 inventories of data and metadata maintained |
| Legal and privacy obligations | PT-2, PT-3, PM-18 Privacy Program Plan | GV.OC-03 legal and regulatory requirements including privacy understood and managed |
| Privacy impact assessment | RA-8 Privacy Impact Assessments | ID.RA |
| Documenting the result | PL-2 System Security and Privacy Plans | GV.PO |
| Data protection following from the category | SC-28 Protection of Information at Rest | PR.DS-01 |
The CSF column is NHI Mgmt Group's indicative mapping; for official crosswalks see the NIST CSF 2.0 Informative References and OLIR Guide.
Common failure patterns
- Categorising the system directly: skipping the information type list produces a single guess that assessors cannot test.
- Forgetting system information: audit logs, configuration and credentials are left out, so a credential store ends up low while the systems it unlocks are high.
- Treating NA as a system value: NOT APPLICABLE is valid only for the confidentiality of an information type, never for a system.
- Using quantity to lower PII impact: SP 800-122 says a small dataset should not by itself set a lower impact level.
- Assuming the privacy baseline follows the impact level: a low system processing sensitive PII still needs privacy controls.
- Never re-categorising: a system acquires new data or integrations over years while its 2018 category stays in the plan.
Practitioner checklist
- Maintain an information type list per system, including logs, configuration, identity data and credentials.
- Use SP 800-60 provisional levels and document every adjustment.
- Express results in FIPS 199 notation for each information type and for the system.
- Apply the high-water mark to select the security baseline.
- Run the SP 800-122 factors with the privacy officer for all PII.
- Decide on the privacy baseline and privacy impact assessment independently of the impact level.
- Rate shared identity services and secrets stores against the highest system they serve.
- Obtain and record authorising official approval of the category in the system plan.
- Trigger re-categorisation on new data types, interconnections or mission changes.
Standards and references
- NIST: FIPS 199, Standards for Security Categorization of Federal Information and Information Systems
- NIST: FIPS 200, Minimum Security Requirements for Federal Information and Information Systems
- NIST: SP 800-60 Vol. 1 Rev. 1, Guide for Mapping Types of Information and Information Systems to Security Categories
- NIST: SP 800-53B, Control Baselines for Information Systems and Organizations
- NIST: SP 800-53 Rev. 5, Security and Privacy Controls for Information Systems and Organizations
- NIST: SP 800-122, Guide to Protecting the Confidentiality of PII
- NIST: SP 800-37 Rev. 2, Risk Management Framework
Related NHI Mgmt Group resources: NIST Risk Management Framework (SP 800-37) Guide · NIST SP 800-53B Control Baselines and Tailoring Guide · NIST SP 800-53 Guide · NIST Privacy Framework 1.1 Guide · FedRAMP Guide