Skip to content
Ayliea

OUR APPROACH

A Rigorous, Repeatable Methodology

Our assessment framework is built on industry-standard security frameworks extended with AI-specific control domains. Every engagement follows the same structured methodology to ensure consistency, completeness, and actionable results.

Built on Established Standards

The Ayliea AI Security Assessment Framework is built on NIST AI Risk Management Framework (AI RMF), NIST Cybersecurity Framework (CSF) 2.0, CIS Controls v8, and ISO/IEC 27001:2022. We extend these with 10 AI-specific control domains that address the unique threat landscape of organizational AI adoption.

Domain IDDomain NameNIST AI RMFNIST CSFCIS v8ISO 27001
AC-1AI Governance & Policy
AC-2AI Asset Management
AC-3Data Protection in AI
AC-4Access Control for AI
AC-5AI Supply Chain Security
AC-6AI Output Validation
AC-7AI Incident Response
AC-8AI Monitoring & Logging
AC-9AI Training & Awareness
AC-10Model Security

The Assessment Lifecycle

Scoping & Kickoff

Phase 1
  • Kickoff meeting with project sponsors
  • Scoping questionnaire review
  • Define assessment boundaries and compliance targets
  • Establish communication channels and schedule

Client involvement: High — key stakeholders participate in kickoff

Discovery

Phase 2
  • AI asset discovery (network, endpoint, interview-based)
  • Data flow mapping and classification
  • Stakeholder interviews across departments
  • Documentation and policy review

Client involvement: Medium — IT and department leads provide access and context

Assessment

Phase 3
  • Control evaluation across 10 AI security domains
  • Compliance gap analysis against selected frameworks
  • Technical testing (API security, access controls, data handling)
  • Evidence collection and documentation

Client involvement: Low — assessment team works independently

Analysis & Reporting

Phase 4
  • Risk scoring using composite methodology
  • Finding prioritization and remediation planning
  • Report drafting (executive summary, technical report, appendices)
  • Quality assurance review

Client involvement: None — internal analysis phase

Delivery & Briefing

Phase 5
  • Executive briefing presentation
  • Technical team walkthrough of findings and remediation roadmap
  • Q&A and clarification sessions
  • Follow-up advisory engagement begins (if included)

Client involvement: High — executive and technical teams attend briefings

Composite Risk Scoring

Risk Score = (Likelihood × Impact × Data Sensitivity) / Control Effectiveness

Each factor is scored 1–4. Higher scores indicate greater risk exposure. Control effectiveness reduces the composite score.

Likelihood

1–4

Probability the threat is exploited, based on exposure and attack surface

Impact

1–4

Business consequence if the risk materializes — financial, reputational, operational

Data Sensitivity

1–4

Classification of data exposed — public, internal, confidential, regulated

Control Effectiveness

1–4

Strength of existing mitigations — divides the composite score to reflect controls

SeverityScore RangeResponse
Critical48 - 64Immediate action required. 48-hour remediation plan.
High24 - 47Priority remediation within 30 days.
Medium8 - 23Planned remediation within 90 days.
Low1 - 7Best practice improvement. Next planning cycle.

See the Methodology in Action

Every assessment starts with a free 30-minute scoping call. We'll show you exactly how this methodology applies to your organization — and what it will cost.