Compliance Reports

Overview#

Guard generates compliance reports that assess your AI deployment against 5 regulatory frameworks used across the GCC and internationally. Reports are generated from actual operational data -- audit logs, policy configurations, PII detection rates, and safety scanning activity -- not self-reported checklists.

Each report includes an overall weighted score, per-requirement assessments with status, evidence of compliance, identified gaps, and actionable recommendations.

Supported Frameworks#

UAE AI Ethics Principles#

Issuing Authority: UAE Ministry of State for Artificial Intelligence Version: 2.0

SDAIA AI Ethics Principles#

Issuing Authority: Saudi Data & AI Authority (SDAIA) Version: 1.0

Qatar National AI Charter#

Issuing Authority: Qatar Ministry of Communications & Information Technology Version: 1.0

Oman AI & Digital Future Program#

Issuing Authority: Oman Ministry of Transport, Communications & IT Version: 2024-2026

EU AI Act#

Issuing Authority: European Union Version: 2024

Generating a Report#

curl -X POST http://localhost:8002/api/v1/compliance/reports/generate \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "framework_id": "uae_ai_ethics",
    "days": 30
  }'

Available framework_id values:

• uae_ai_ethics
• sdaia_ai_ethics
• qatar_ai_charter
• oman_ai_digital
• eu_ai_act

The days parameter controls the assessment period (1-365 days, default 30).

Report Response#

{
  "id": "r-a1b2c3d4-...",
  "framework_id": "uae_ai_ethics",
  "framework_name": "UAE AI Ethics Principles",
  "version": "2.0",
  "issuing_authority": "UAE Ministry of State for Artificial Intelligence",
  "period_start": "2026-01-17T00:00:00Z",
  "period_end": "2026-02-16T00:00:00Z",
  "overall_score": 85.7,
  "executive_summary": "Compliance assessment against UAE AI Ethics Principles yielded an overall score of 85.7%. Of 7 requirements assessed: 5 compliant, 1 partially compliant, 1 non-compliant. Analysis covers 1,247 AI interactions with 12 policy violations and 3 PII detections. Priority areas requiring attention: Sustainability.",
  "sections": [
    {
      "requirement_id": "uae_transparency",
      "requirement_name": "Transparency",
      "description": "AI systems must provide clear explanations of how decisions are made and what data is used",
      "weight": 0.15,
      "status": "compliant",
      "evidence": ["audit_trail_active", "policy_engine_active"],
      "gaps": [],
      "recommendations": []
    }
  ],
  "generated_by": "admin@example.com",
  "created_at": "2026-02-16T10:00:00Z"
}

How Scoring Works#

Audit Signals#

Guard gathers operational signals from your deployment data to assess each requirement:

Per-Requirement Status#

Each requirement maps to a set of audit signals. The status is determined by the ratio of passing signals:

Overall Score Calculation#

The overall score is a weighted average across all requirements:

• Compliant = 100 points
• Partial = 50 points
• Non-Compliant = 0 points
• Not Assessed = 0 points

Each requirement's score is multiplied by its weight, and the total is normalized to a percentage.

Listing Frameworks#

curl http://localhost:8002/api/v1/compliance/reports/frameworks \
  -H "Authorization: Bearer $TOKEN"

[
  {
    "id": "uae_ai_ethics",
    "name": "UAE AI Ethics Principles",
    "version": "2.0",
    "issuing_authority": "UAE Ministry of State for Artificial Intelligence",
    "requirement_count": 7
  },
  {
    "id": "sdaia_ai_ethics",
    "name": "SDAIA AI Ethics Principles",
    "version": "1.0",
    "issuing_authority": "Saudi Data & AI Authority (SDAIA)",
    "requirement_count": 6
  }
]

Report History and Comparison#

Listing Reports#

curl "http://localhost:8002/api/v1/compliance/reports?framework_id=uae_ai_ethics&limit=10" \
  -H "Authorization: Bearer $TOKEN"

Viewing a Report#

Exporting a Report#

Comparing Reports#

curl "http://localhost:8002/api/v1/compliance/reports/$REPORT_A/diff/$REPORT_B" \
  -H "Authorization: Bearer $TOKEN"

{
  "report_a_id": "...",
  "report_b_id": "...",
  "framework_id": "uae_ai_ethics",
  "score_change": 12.5,
  "section_changes": [
    {
      "requirement_id": "uae_privacy",
      "requirement_name": "Privacy",
      "status_before": "non_compliant",
      "status_after": "compliant",
      "improved": true
    }
  ]
}