Compliance & Security

Enterprise security. Regulatory compliance. Human oversight.

Built for SAMA-regulated institutions. PDPL-compliant. Every AI output verified by a human before delivery.

app.abwab.ai/compliance

Human-Verified AI

QA before delivery
1

AI drafts

extract, score, flag

2

Human QA

review & correct

3

Delivered

logged + auditable

Every AI output passes human QA before it ships.

Regulatory compliance

Built for Saudi Arabia's regulatory environment.

Abwab maps to the SAMA Cybersecurity Framework, PDPL data protection, SAMA outsourcing guidelines, and Sharia compliance, with model governance built in.

  • SAMA Cybersecurity Framework, all four domains
  • PDPL-compliant for every borrower record
  • Sharia-validated, used by Kafalah

Compliance coverage

SAMA Cybersecurity Framework
PDPL data protection
SAMA Outsourcing guidelines
Sharia-validated
Model governance & explainability

Tracked with automated coverage scoring.

Data security

Enterprise-grade security at every layer.

Encryption in transit and at rest, granular role-based access, regular third-party penetration testing, and in-Kingdom data residency.

  • TLS 1.3 in transit, AES-256 at rest
  • Granular role-based access control
  • On-premise option for full data sovereignty

Security

Encryption

TLS 1.3 in transit · AES-256 at rest

Access control

granular role-based access

Penetration testing

regular, third-party

Data residency

in-Kingdom · on-prem option

Enterprise-grade at every layer.

Audit trail

Every decision, every actor, every timestamp.

A complete, immutable audit trail for every credit decision: who reviewed it, what data was used, which policies applied, and what the AI recommended.

  • Immutable, exportable logs
  • Full inputs, paths, and outputs
  • Satisfies regulatory and internal review

Decision log

exportable
App #4821 · APPROVEmodel v12
Override · refer → approveanalyst S.A.
Policy update · v11 → v12risk team

Full inputs, paths, and outputs. Version controlled for audit.

Human in the loop

Human-Verified AI

Every AI output passes through human QA before delivery. Our annotation team reviews, validates, and corrects every assessment. No black box. No unverified automation.

1

AI generates

Models extract data, calculate scores, flag risks, and draft the credit assessment with full reasoning chains.

2

Humans verify

Trained analysts review every output, checking data accuracy, scoring logic, and narrative coherence before anything ships.

3

Full audit trail

Every decision is logged: what the AI produced, what the human changed, and why. Complete transparency for regulators and auditors.

Frequently asked questions

Yes. Abwab is built to align with the SAMA Cybersecurity Framework across all four domains: Leadership & Governance, Risk Management & Compliance, Operations & Technology, and Third Party Cybersecurity. We also comply with SAMA outsourcing guidelines.

Review our security and compliance posture.