---
title: "Artificial Intelligence Guideline (QCB AI Guideline)"
url: "https://grc.plus/regulation/qcb-artificial-intelligence-guideline/"
type: "Regulation"
updated: "2026-11-19"
publisher: "GRC+"
---

# Artificial Intelligence Guideline (QCB AI Guideline)

The Guideline applies when a QCB regulated entity develops and implements AI on its own, purchases AI systems, or outsources processes or functions that rely directly on AI (section 5).

Governance: an AI strategy reviewed periodically (section 6), corporate governance (section 7), an AI governance policy (section 8) and an AI risk management framework (section 9). An entity must classify an AI system as high-risk where it can harm natural persons, such as decisions on consumer access to financial services, material decisions about employees, or processing sensitive personal information (section 9.7).

Register and approval: an entity keeps a register of all its AI systems, highlighting high-risk systems and the provider, and discloses the full register to QCB every year and on request (section 10). It needs QCB approval before launching a new AI system as a provider, before material changes, and before signing a high-risk AI purchase, licensing or outsourcing agreement (section 11). Outsourcing any AI function needs documented Board approval, and QCB expects its prior consent (section 12). The full training, validation and testing results of a new high-risk AI system go to QCB for approval (section 14).

The Guideline also covers human oversight, training and testing approvals, AI data governance, high-risk AI management, AI security, documentation and monitoring, and customer information, consent, recourse and opt-out (sections 13 to 22). Exemptions need QCB approval (section 23). Section 24 requires entities operating in Qatar to comply, besides the QCB Law and the Guideline, with the Sector-Specific Security Regulation, Law No. 13 of 2016 on Personal Data Privacy Protection, the QCB Technology Risks Circular of January 2018, the QCB Cloud Computing Regulation 2024 when they use cloud-based AI, and other QCB regulations and guidelines.

**Ahmet Ucar**Principal Consultant

- [Book a call](https://meetings-eu1.hubspot.com/ahmet-ucar)
- [+31 20 241 66 00](tel:+31202416600)
- [sales@grc.plus](mailto:sales@grc.plus)

## Review Info

- Last Update: 19/11/2026

Do you want to receive updates on QCB AI Guideline?

No spam. No advertising. Regulatory updates only.

[Subscribe to updates](https://meetings-eu1.hubspot.com/ahmet-ucar)

## Quick Facts on QCB AI Guideline

## Reference

- Official reference: QA-QCB-AI

[Read the official text](https://www.qcb.gov.qa/Documents/BankInstructions2024/AR/annex160.pdf)

## Purpose

It sets requirements and guidance for the use of AI systems by all entities regulated by QCB, so that their use is safe, secure and efficient, and defines principles for using AI, assessing its risks and categorising AI systems (section 4).

## Jurisdiction & Supervisory Authorities

- Supervisory authorities: [Qatar Central Bank](https://grc.plus/authority/qa-qatar-central-bank/)

## Key Requirements

### AI strategy

Define an AI strategy based on the entity's needs and risk appetite and review it periodically.

### AI governance policy

Put in place corporate governance and an AI governance policy.

### AI risk management

Assess and rate AI risk, and classify systems that can harm natural persons as high-risk.

### AI register

Keep a register of all AI systems, highlighting high-risk systems and the provider, and disclose it to QCB annually and on request; disclose the high-risk criteria used and a high-level risk and impact assessment.

### QCB approval

Obtain QCB approval before launching or materially changing an AI system as a provider, and before signing a high-risk AI agreement.

### Outsourcing of AI

Obtain Board of Directors approval, documented, before outsourcing any AI function; carry out due diligence before selecting the provider, review it periodically, keep a service agreement with minimum clauses and a contingency plan. QCB expects its prior consent and regular due diligence on the provider.

### Human oversight

Apply human oversight to AI in use.

### Training and testing results to QCB

Submit the full training, validation and testing results of any new high-risk AI system to QCB for approval; as a user, submit the provider's instructions and results and your own test results, data sources and human oversight plan.

### Customer information and recourse

Offer customers a way to ask about AI decisions and to have a negative decision reviewed by a qualified human decision-maker (section 21). QCB also expects customer information, consent to the risks of AI and a considered choice on an opt-out, with another recourse channel where no opt-out is offered (sections 20 and 22).

## Related

Regulation

### [Technology Risks Regulation (QCB Technology Risks)](https://grc.plus/regulation/qcb-technology-risks-regulation-banks/)

Sets cyber security requirements for banks in Qatar, with one-hour reports of critical incidents and an annual board-signed compliance report.

Regulation

### [Cloud Computing Regulation (QCB Cloud Regulation)](https://grc.plus/regulation/qcb-cloud-computing-regulation/)

Requires QCB-regulated entities to get approval for material cloud arrangements and to process personal and financial data in Qatar.

Regulation

### [ISO/IEC 42001: Artificial Intelligence Management System](https://grc.plus/regulation/iso-iec-42001-ai-management-system/)

Specifies a certifiable management system for organisations that develop, provide or use AI to govern it and assess its risks and impacts.

## Talk to a Principal Consultant

Thirty minutes, no slides. You leave knowing which requirements apply to you and what it takes to meet them.

[Book a Scoping Call](https://meetings-eu1.hubspot.com/ahmet-ucar)
