ConsultingWorkbench-backed AI security engagements — map, attack, defend, and prove your AI systems.
Scope a Review
AI Product Security Assessment — Evidence Packs

Evidence Pack System

Structured outputs from AIPSA assessment and adversarial testing. Auditable, exportable, and ready for customers, auditors, and boards.

Start with the Scorecard

Evidence Pack Types

AIPSA Assessed

Organization completed the AIPSA scorecard assessment. Includes domain scores, maturity level, and gap summary.

  • Domain score matrix
  • Maturity level determination
  • Gap summary by domain group
  • Assessment date and scope

AIPSA Tested

Adversarial testing completed across scoped AI systems. Includes test scope, methodology, and findings summary.

  • Test scope and systems covered
  • Methodology reference
  • Findings summary (no exploits)
  • Retest status

AIPSA Evidence Pack

Full evidence bundle from a scoped engagement. Combines assessment, testing, and control mapping.

  • Scorecard results
  • Test findings
  • Control mapping to AIPSA domains
  • Executive summary for customers

AIPSA Verified Controls

Specific controls verified as implemented during an evidence review. Scoped to named domains.

  • Named controls verified
  • Domain scope
  • Evidence artifacts referenced
  • Verifier attestation

Common Use Cases

  • Customer security questionnaire responses
  • SOC 2, ISO 27001, or regulatory audit support
  • Board and executive AI security reporting
  • Vendor due diligence packages
  • AI security trust center content

Domain Scope

Evidence packs can be scoped to any combination of AIPSA domain groups.

Inventory & Architecture

3 domains

Adversarial Testing

3 domains

RAG & Data Authorization

3 domains

Agentic Permissions

3 domains

Detection & Incident Response

3 domains

Governance Evidence

3 domains

Generate evidence for your program

Start with the AIPSA scorecard to establish your baseline. Evidence packs are generated as part of assessment and consulting engagements.