Framework · Ethics & Artificial Intelligence

Niyyah–Amānah AI Ethics Framework (NAAF)

An Islamic Framework for Responsible and Ethical Use of Artificial Intelligence

Purpose

NAAF is an Islamic ethical self-examination framework designed to help individuals assess whether their use of Artificial Intelligence is responsible, honest, intellectually accountable and consistent with Islamic ethical principles.

It does not merely ask whether AI use is technically possible or legally permissible. It encourages the user to examine why AI is being used, whether its use is genuinely necessary, whether human responsibility is being retained, whether its outputs have been verified, and whether the resulting work can be honestly defended before Allah.

Imam Abdul-Azeez Onike Morufu

Framework type
Ethical self-examination
Gates
7 sequential ethical gates
Response levels
Green · Amber · Red
Access
Open access
Language
English
Ecosystem
The Onike Framework™
The Seven Ethical Gates

From Niyyah to Amānah

Each gate is a pause. At every stage, the answer is YES, UNCERTAIN or NO — and the framework responds accordingly.

Gate 1

Niyyah

Why am I using AI?

Niyyah (intention) provides the moral starting point for human action.

Warning sign

I am using AI because it allows me to obtain the result without doing work that I am actually responsible for doing.

Self-examination questions
  • What exactly am I trying to accomplish?
  • Is my intention constructive and legitimate?
  • Am I using AI to learn, assist, organise, analyse or create?
  • Would I be comfortable stating my actual intention to a trusted scholar or colleague?
Gate 2

Need

Do I genuinely need it, or am I merely avoiding effort?

Islam values responsible effort, diligence and fulfilment of obligations.

Warning sign

AI is doing this mainly because I do not want to make the effort myself.

Self-examination questions
  • Do I genuinely need AI for this task?
  • What part of the task could I reasonably do myself?
  • Does the task require my own reasoning, reflection, skill or personal experience?
  • Would using AI undermine the purpose of the task?
Gate 3

Alternative

What legitimate human or scholarly sources are available?

Knowledge carries responsibility. AI should not automatically replace qualified scholarship or primary sources.

Warning sign

I am accepting the AI answer even though I know I should consult the original or authoritative source.

Self-examination questions
  • Is there a Qur'anic, Prophetic, scholarly, academic or professional source I should consult directly?
  • Is AI being used as a supplementary tool rather than replacing an authoritative source?
  • Have I consulted the original source where accuracy matters?
Gate 4

Responsibility

Am I retaining intellectual responsibility for the work?

Human accountability remains central. Delegating a task to AI does not necessarily delegate the ethical responsibility.

Warning sign

I do not fully understand or stand behind this work, but I intend to present it as my own.

Self-examination questions
  • Do I understand the material I am submitting?
  • Can I explain and defend the important claims?
  • Have I contributed meaningful human judgement?
  • Would I accept responsibility if an important part of the AI-generated output proved wrong?
Gate 5

Verification

Have I checked what AI has given me?

Responsible use of information requires care concerning truthfulness, accuracy and reliability.

Warning sign

I have not checked it, but I intend to present it as factual.

Self-examination questions
  • Have I checked important factual claims?
  • Have I verified quotations and references against the original sources?
  • Have I checked Qur'anic verses, hadith, scholarly opinions and legal claims against reliable sources?
  • Have I distinguished between established knowledge and AI-generated inference?
Gate 6

Transparency

Would my use create a misleading impression if undisclosed?

Transparency protects trust and reduces the risk of deception or misrepresentation.

Warning sign

If people knew how much AI contributed, they would understand my work differently.

Self-examination questions
  • Would a reasonable reader, assessor, employer, client or audience misunderstand my contribution if I did not disclose the AI assistance?
  • Does the institution, journal, employer or assessment provider require disclosure?
  • Am I allowing people to believe that I performed work that was substantially performed by AI?
Gate 7

Amānah

Can I honestly stand behind this work before Allah?

Amānah (trustworthiness/trust) represents the culmination of the framework.

Warning sign

I would not be comfortable explaining my use of AI before Allah.

Self-examination questions
  • Can I honestly stand behind this work?
  • Have I fulfilled the responsibility entrusted to me?
  • Have I knowingly allowed falsehood, deception or misrepresentation to enter the work?
  • Would I be comfortable explaining my use of AI before Allah?
Decision Path

The NAAF pathway

Gate 1

Niyyah

Gate 2

Need

Gate 3

Alternative

Gate 4

Responsibility

Gate 5

Verification

Gate 6

Transparency

Gate 7

Amānah

YES

Proceed to the next gate.

UNCERTAIN

Pause, investigate, consult or modify.

NO

Stop, reconsider or change the use of AI.

The framework functions as an ethical pause mechanism rather than simply a scoring instrument.

Response Levels

Three-level ethical response

Green

Responsible Use

The user has a legitimate purpose, retains responsibility, verifies relevant information and does not create a misleading impression. Proceed, while maintaining appropriate human judgement.

Amber

Review Required

There is uncertainty about necessity, authorship, verification, disclosure or responsibility. Pause. Investigate, verify, consult an appropriate source or modify the use of AI.

Red

Ethical Concern

The user intends to deceive, misrepresent authorship, avoid a responsibility that should be performed personally, knowingly present unverified false information, or otherwise use AI inconsistently with responsible conduct. Do not proceed in the current form.

Central Principle

AI may assist, but it does not replace moral responsibility

AI may assist the human being, but it does not replace the human being's moral responsibility.

The framework distinguishes between AI assistance and AI substitution of responsibility. The more consequential the task, the greater the need for human judgement, verification, accountability and transparency.

Scope

Where NAAF applies

Academic research and writing
Islamic scholarship and teaching
Professional work
Institutional decision-making
Content creation
Islamic finance
Education
Research and publication
Religious communication
Policy development
Business and organisational use of AI

NAAF is intended as an ethical reflection and governance framework, not as a substitute for qualified Islamic legal advice, institutional policy or applicable law.

Core Proposition

One sentence summary

Before using AI, examine your Niyyah; establish your Need; consider legitimate Alternatives; retain Responsibility; perform Verification; practise appropriate Transparency; and finally ask whether your use constitutes an Amānah you can honestly stand behind before Allah.

Begin

Put the framework into practice

Use the interactive self-assessment to walk through the seven gates for a specific AI-assisted task. It takes under five minutes and produces a Green, Amber or Red response with practical next steps.

Cite this framework

How to cite this framework

Abdul-Azeez, O. M. (2026). Niyyah–Amānah AI Ethics Framework (NAAF): An Islamic framework for responsible and ethical use of artificial intelligence. The Onike Framework. https://onikeframework.org/naaf#the-seven-gates

Peer Review & Continuous Improvement

Help Improve The Onike Framework™

The Onike Framework™ is committed to continuous learning, evidence-based improvement, and collaborative knowledge creation. Your feedback, professional experience, and constructive recommendations are invaluable in refining our frameworks and expanding their global relevance.

Peer Reviewed
Practitioner Reviewed
Field Tested
Recently Updated