HITRUST AI Security Assessment and Certification Specification

Restrict access to data used for AI

HITRUST CSF requirement statement [?] (11.01cAISecSystem.6)

The organization restricts all access to the data used to 
(1) train, test, and validate AI models; 
(2) fine-tune AI models; and 
enhance AI prompts via RAG (both the
(3) original and
(4) vectorized formats stored as embeddings, if used)
following the least privilege principle. 
This access is controlled in accordance with the organization’s policies regarding 
(5) access management (including approvals, revocations, periodic access reviews), and 
(6) authentication (which calls for multi-factor authentication or a similar level of protection).

Evaluative elements in this requirement statement [?]
1. The organization restricts all access to the data used to train, test, and validate AI 
models following the least privilege principle.
2. The organization restricts all access to the data used to fine-tune AI models following 
the least privilege principle.
3. The organization restricts all access to the original (non-vectorized) data used to enhance AI prompts
via RAG following the least privilege principle, if applicable.
4. The organization restricts all access to the embeddings data used to enhance AI prompts via RAG
following the least privilege principle, if applicable.
5. This access is controlled in accordance with the organization’s policies regarding 
access management (including approvals, revocations, periodic access reviews).
6. This access is controlled in accordance with the organization’s policies regarding 
authentication (which calls for multi-factor authentication or a similar level of protection).


Illustrative procedures for use during assessments [?]

  • Policy: Examine policies related to each evaluative element within the requirement statement. Validate the existence of a written or undocumented policy as defined in the HITRUST scoring rubric.

  • Procedure: Examine evidence that written or undocumented procedures exist as defined in the HITRUST scoring rubric. Determine if the procedures and address the operational aspects of how to perform each evaluative element within the requirement statement.

  • Implemented: Examine evidence that all evaluative elements within the requirement statement have been implemented as defined in the HITRUST scoring rubric, using a sample based test where possible for each evaluative element. Example test(s):
    • For example, review the AI system to ensure The organization restricts all access to the data used to train, test, and validate AI models; tune AI models; and enhance AI prompts via RAG, if applicable, following the least privilege principle. Further, confirm this access is controlled in accordance with the organization’s policies regarding access management (including approvals, revocations, periodic access reviews) and authentication.

  • Measured: Examine measurements that formally evaluate and communicate the operation and/or performance of each evaluative element within the requirement statement. Determine the percentage of evaluative elements addressed by the organization’s operational and/or independent measure(s) or metric(s) as defined in the HITRUST scoring rubric. Determine if the measurements include independent and/or operational measure(s) or metric(s) as defined in the HITRUST scoring rubric. Example test(s):
    • For example, measures indicate if the organization restricts all access to the data used to train, test, and validate AI models; tune AI models; and enhance AI prompts via RAG, if applicable, following the least privilege principle. Reviews, tests, or audits are completed by the organization to measure the effectiveness of the implemented controls and to confirm that all access is controlled in accordance with the organization’s policies regarding access management (including approvals, revocations, periodic access reviews) and authentication.

  • Managed: Examine evidence that a written or undocumented risk treatment process exists, as defined in the HITRUST scoring rubric. Determine the frequency that the risk treatment process was applied to issues identified for each evaluative element within the requirement statement.

Placement of this requirement in the HITRUST CSF [?]

  • Assessment domain: 11 Access Control
  • Control category: 01.0 – Access Control
  • Control reference: 01.c – Privilege Management

Specific to which parts of the overall AI system? [?]
AI application layer:
  • Prompt enhancement via RAG, and associated RAG data sources
AI platform layer:
  • Model tuning and associated datasets
  • AI datasets and data pipelines


Discussed in which authoritative AI security sources? [?]
  • LLM AI Cybersecurity & Governance Checklist
    Feb. 2024, © The OWASP Foundation
    • Where:
      • 3. Checklist > 3.9. Using or implementing large language model solutions > Bullet #2
      • 3. Checklist > 3.9. Using or implementing large language model solutions > Bullet #4

  • Securing Machine Learning Algorithms
    2021, © European Union Agency for Cybersecurity (ENISA)
    • Where:
      • 4.1- Security Controls > Organizational > Apply a RBAC model, respecting the least privilege principle
      • 4.1- Security Controls > Technical > Ensure appropriate protection is deployed for test environments

Discussed in which commercial AI security sources? [?]
  • Databricks AI Security Framework
    Sept. 2024, © Databricks
    • Where:
      • Control DASF 1: SSO with IdP and MFA
      • Control DASF 2: Sync users and groups
      • Control DASF 5: Control access to data and other objects
      • Control DASF 16: Secure model features
      • Control DASF 43: Use access control lists
      • Control DASF 57: Use attribute-based access controls (ABAC)

  • Google Secure AI Framework
    June 2023, © Google
    • Where:
      • Step 4. Apply the six core elements of the SAIF > Expand strong security foundations to the AI ecosystem > Prepare to store and track supply chain assets, code, and training data

  • HiddenLayer’s 2024 AI Threat Landscape Report
    2024, © HiddenLayer
    • Where:
      • Part 4: Predictions and recommendations > 3. Data security and privacy > Bullet #1

  • Snowflake AI Security Framework
    2024, © Snowflake Inc.
    • Where:
      • Training data leakage > Mitigations > Access controls

Control functions against which AI security threats? [?]
Additional information
  • Q: When will this requirement included in an assessment? [?]
    • This requirement is included when the assessment’s in-scope AI system(s) leverage data-driven AI models (e.g., non-generative machine learning models, generative AI models).
    • The Security for AI systems regulatory factor must also be present in the assessment.

  • Q: Will this requirement be externally inheritable? [?] [?]
    • Yes, fully. This requirement may be the sole responsibility of the AI model creator. Or, depending on the AI system’s architecture, only evaluative elements that are the sole responsibility of the AI model creator apply.