ISLAMIC CODE FOR THE APPLICATION OF ARTIFICIAL INTELLIGENCE
The Almighty Allah is the sole Creator and Sovereign of all worlds. To Him alone belongs all Praise. O Allah! Bless Muhammad, his family, and his community!
Man has been endowed with the faculty of understanding and free will so that he may consciously worship the Lord, fulfill His commandments, and abstain from the forbidden. All human actions, including scientific and technological developments, are to be evaluated through the lens of compliance with Divine laws.
In Islam, the human being is granted a special status as vicegerent (khalifah) on earth: he is entrusted with the task of improving the world, increasing goodness, and preventing evil. Technological progress, including the development of artificial intelligence, is regarded as part of this divine gift of knowledge (‘ilm). However, the responsibility placed upon man requires that he utilize new opportunities in accordance with Shari‘ah and the principles of akhlaq, taking care to preserve faith, life, intellect, property, and honor.
The present Code is intended to demonstrate how these supreme Islamic guidelines may be applied to the sphere of artificial intelligence. It describes how developers, religious institutions, and all interested parties may utilize AI for the benefit of society, without transgressing the limits set by Shari‘ah or violating moral values.
The fundamental principle of this Code is the primacy of religious norms over technological innovations: any scientific and technical achievements are to be evaluated solely through the prism of their compliance with the prescriptions of Islam. Technological progress must be aligned with Islamic values, and not vice versa. In cases where technologies present potential risks to faith, life, intellect, property, or honor, their use must be restricted or suspended until such risks are eliminated.
Table of contents
1.1. Purpose and Objectives of the Code. 3
1.2. Basis and Scope of Application. 3
1.3. “Definitions” (Glossary) 3
1.4. Principle of Voluntariness and Recognition. 4
2. Islamic Principles and Core Values. 4
2.1. System of Mandatory Provisions Based on the Principles of Maturidi Creed. 4
2.2. System of Mandatory Provisions Based on the Principles of Hanafi Fiqh. 6
2.3. System of Mandatory Provisions Based on the Principles of Islamic Akhlaq. 11
3. Requirements for Design and Implementation. 13
3.1. Design and Research Stage. 13
3.2. Implementation and Operation Stage. 13
3.3. Update and Control Stage. 14
4. Organization of Islamic Oversight 15
4.1. Authorized Persons for Islamic Norms in AI. 15
4.2. Commissions on Islamic AI Norms. 15
4.3. Register of Islamic AI Solutions. 16
5. Measures for Compliance and Development of the Code. 17
5.1. Obligations of Participants. 17
5.2. Education and Awareness. 17
5.3. Experience Sharing and Cooperation. 17
5.4. Revision and Updating of the Code. 18
6.1. Harmony Between Progress and Shari‘ah Norms. 18
6.2. Strengthening Trust and Improving Society. 19
6.3. Call for Shared Responsibility. 19
6.4. The Future of “Strong AI” and Transhumanism.. 20
1.1. Purpose and Objectives of the Code
1.1.1. Purpose of the Code. The purpose of this Code is to establish a normative framework for the application of artificial intelligence (AI) technologies in the Islamic context.
1.1.2. Main Objectives:
• Risk Prevention: Minimization of potential ethical, social, and religious risks in the use of AI, with mandatory assessment of each innovation through the prism of Shari‘ah.
• Clarification of Responsibility: Clear definition of the roles and duties of participants in the life cycle of artificial intelligence systems.
• Establishment of Principles: Formation of behavioral norms whereby technological progress is subordinated to Islamic values and religious norms.
The life cycle refers to the stages of research, design, development, testing, implementation, operation, and improvement of AI technologies.
1.2. Basis and Scope of Application
1.2.1. Shari‘ah Foundation. The Code is based on Shari‘ah (Qur’an, Sunnah, ijma‘, qiyas), the fundamentals of creed (‘aqidah), and the values of Islamic ethical teaching (akhlaq), and also takes into account the experience of existing secular codes of AI ethics and legal norms in the field of digital technologies.
1.2.2. Intended Audience of the Code. This document is intended for developers, religious organizations, Islamic educational institutions, and all those involved in the design, implementation, and use of AI in societies where the principles of Islam are observed.
1.2.3. Civil Applications. The recommendations of the Code cover non-military scenarios for the use of AI technologies, including (but not limited to):
• Chatbots, recommender systems, and platforms for text content generation,
• Islamic banking services,
• Educational platforms with integrated generation of graphic, audio, and video materials,
• Telemedicine solutions,
• Systems for the analysis, generation, and certification of halal products,
• Multimodal AI systems creating text, graphics, audio, video, software code, and analytical data for various applications.
The Code may also be applied to research and academic projects, even if their results are not directly made public. Early consideration of Shari‘ah and ethical standards during data collection and model training helps to avoid potential issues at the implementation stage.
1.3.1. “Artificial Intelligence System (AIS)” — any technology utilizing machine learning, neural networks, big data analysis, and other methods, including systems capable of generating multimodal content: texts, graphic images (illustrations, 3D models, UI/UX design), audio materials (music, voiceovers, podcasts, sounds), video materials (animation, editing, deepfakes), software code, and analytical data.
1.3.2. “AI Actors” — a collective designation for the participants in the AI life cycle:
• Developers (engineers, programmers, etc.),
• Integrators (organizations or individuals implementing AI into existing processes),
• Users (individuals, companies, institutions directly utilizing AI),
• Religious Experts (muftis, Shari‘ah councils, and other specialists authorized to assess compliance with Islamic norms).
1.3.3. “Islamic Context” — the aggregate of norms, values, and principles based on Shari‘ah, akhlaq, and ‘aqidah that define the permissible framework for the development and application of artificial intelligence technologies in Islamic society.
1.3.4. “Maqasid al-Shari‘ah” — the higher objectives of Shari‘ah, which include the preservation of religion, life, intellect, property, and honor.
1.3.5. “Islamic AI” — artificial intelligence developed and operated in strict accordance with Shari‘ah norms and principles, excluding haram functionality and contributing to the spiritual development of the user. It is oriented toward honesty, justice, strengthening of iman, and the maximum benefit to society, without substituting for the free will of the human being. At the same time, it is recognized that all moral responsibility lies with the human, not with algorithms.
1.4. Principle of Voluntariness and Recognition
1.4.1. Voluntary Nature. AI actors adopt this Code on a voluntary basis, reflecting their commitment to comply with Islamic norms and generally accepted standards for the responsible use of technology.
1.4.2. Responsibility. AI cannot assume moral or legal responsibility for decisions, which must remain with the human being.
This is based on the principle that only a human (mukallaf) is endowed with intellect (‘aql) and free will, and therefore is accountable before Allah.
1.4.3. Consideration by Authorities. Adherence to the Code may be taken into account by religious and state authorities when assessing the reliability and legitimacy of technological products and services.
2. Islamic Principles and Core Values
2.1. System of Mandatory Provisions Based on the Principles of Maturidi Creed
2.1.1. Tawhid and Warning Against Techno-Deification
The development and operation of AI must be founded upon strict observance of the principle of monotheism (tawhid). It is impermissible to regard AI as an independent force or to attribute to it divine qualities, as well as to idealize technology as an alternative to religious values.
2.1.2. Predestination Without Fatalism
Divine predestination is recognized, yet human responsibility for choice and actions in the use of AI remains. It is prohibited to justify errors and abuses by referencing the inevitability or autonomy of technologies.
2.1.3. Free Will and Personal Responsibility
Each person is granted free will and bears full responsibility for decisions related to the use of AI. It is impermissible to shift blame or ultimate decisions to automated systems.
2.1.4. The Role of Reason Alongside Revelation
Decisions in the field of AI must be based on a combination of rational analysis and religious prescriptions. No technological innovation may contradict the foundations of belief.
2.1.5. The Nature of Knowledge and Epistemological Criteria
It is prohibited to train AI on materials containing forbidden or misleading content (such as slander, promotion of shirk/kufr).
2.1.6. Justice and the Principle of Moderation
AI must be implemented without violating justice or infringing upon anyone’s rights. The imposition of excessive technical requirements that exceed the real capacities of individuals and society is not permissible.
2.1.7. Principle of Ma‘ruf and Munkar
All processes related to AI must be directed toward promoting public good and preventing evil. AI may be used exclusively for righteous and socially beneficial purposes.
2.1.8. Source Criticality and Verification of Authenticity
Any information used in AI is subject to mandatory verification. Algorithms must provide for filtering of unreliable and inauthentic data.
2.1.9. Wisdom and Precaution
The implementation of AI must be accompanied by thorough assessment of risks and long-term consequences. Hasty implementation that may harm religious or social values is not permitted.
2.1.10. Reliance Upon Allah
The use of AI must be accompanied by the awareness of dependence on the will of the Almighty. It is impermissible to rely on technology as an absolute solution or as a substitute for divine decree.
2.1.11. Duty to Seek Truth Through Reason
A Muslim is obliged to seek knowledge and rational comprehension of data, including in work with AI. The use of AI should contribute to the development of competencies, but not substitute for personal pursuit of truth and critical thinking.
2.2. System of Mandatory Provisions Based on the Principles of Hanafi Fiqh
I. General Legal Postulates and Principles
2.2.1. Principle of Intention
Any action in the field of AI must be performed with pure and conscious intention, in accordance with the objectives of Shari‘ah. The permissibility is evaluated based on the true motive and aim of the developer, operator, or user. Absence of sincerity renders the action impermissible and devoid of reward.
2.2.2. Fundamental Principle of Permissibility
The use of technology is permitted by default, unless there is a direct Shari‘ah prohibition. The burden of proof for prohibition rests upon the party asserting the restriction. All doubts are construed in favor of permissibility.
2.2.3. Principle of Certainty
All decisions and processes related to AI must be based on reliable and verified information. Previously established permissions may not be revoked solely on the basis of doubt, except in cases where serious and substantiated doubts arise. A stable situation may only be revised upon the emergence of reliable evidence.
2.2.4. Principle of Harm Prevention
It is prohibited to create, implement, or operate AI systems that cause harm to a person, property, religion, or society. Any identified harm must be immediately eliminated. The responsibility for preventing harm rests with all participants in the system’s life cycle.
2.2.5. Principle of Facilitation in Difficulty
In cases of objective difficulty, simplified procedures may be applied, provided that this does not contravene Shari‘ah prohibitions. The goal is to reduce unjustified burdens for developers and users. Facilitation is applied strictly within the bounds of permissibility.
2.2.6. Principle of Necessity
In cases of dire necessity, temporary departure from ordinary prohibitions is permitted in the absence of alternatives. Such measures are allowed only to the extent required to avert the threat. Once necessity ceases, the application of dispensations is discontinued.
2.2.7. Principle of Forgiveness of Error and Coercion
Responsibility for unintentional errors, forgetfulness, or actions under duress is mitigated or removed. Establishing culpability requires deliberate intent. Automated systems must account for such circumstances in evaluating human behavior.
2.2.8. Principle of Justice
Justice must be ensured at all stages of the AI life cycle, including design, implementation, and operation. Any form of discrimination or bias in algorithms and outcomes is prohibited. Justice is to be assessed regularly and systematically.
2.2.9. Principle of Responsibility
Developers, owners, and operators of AI bear full responsibility for all consequences of system operation, including potential harm. Responsibility covers both technical and moral aspects of AI application. Compensation measures for harm must be provided in advance.
2.2.10. Principle of Means-End Consistency
Means employed to achieve objectives must be evaluated in light of their ultimate results. Anything that leads to the forbidden becomes impermissible, even if originally allowed. Actions must be aligned with intentions and must not contradict the purposes of Shari‘ah.
2.2.11. Principle of Preventing Means to Sin
It is impermissible to create or disseminate technologies that are highly likely to lead to the commission of forbidden acts. Any system facilitating the violation of moral or legal norms must be blocked or eliminated. Responsibility for preventing negative consequences rests with all participants in the process.
2.2.12. Principle of Prioritizing Benefit Over Harm
The use of AI is allowed only in cases where public and individual benefit clearly outweighs possible harm. Benefit-harm assessment must be carried out at all stages of system deployment. In cases of doubt, priority is given to societal interests.
2.2.13. Principle of Istihsan (Preference for the Best Solution)
In the event of legal conflicts or unresolved issues, preference is given to the most just and humane solution. Both societal and individual interests are assessed. Flexibility in decision-making is encouraged within Shari‘ah norms.
2.2.14. Principle of Analogy (Qiyas)
Legal decisions regarding new issues in the field of AI are made on the basis of analogy with established Shari‘ah norms and precedents. Strict conformity of meaning and purpose is required. Artificial extension of analogy beyond what is permitted is impermissible.
2.2.15. Principle of Necessity of Required Means
Whatever is necessary to fulfill an obligatory requirement also becomes obligatory. This applies to both technical and organizational measures to ensure compliance with Shari‘ah prescriptions. Failure to observe auxiliary measures results in violation of the primary duty.
2.2.16. Principle of “Income With Responsibility”
A person deriving income from the application of AI bears financial responsibility for any damage caused by their activity. Profit may not be obtained without readiness to compensate for harm. This rule is binding on all commercial participants in the AI market.
2.2.17. Duty of Continuous Monitoring
Continuous technical and organizational oversight over AI operation must be ensured, including regular audits and compliance checks. Any violations detected are subject to immediate rectification.
2.2.18. Principle of Preventing Harm Prevails Over Benefit
Priority is given to measures that minimize and eliminate harm, even if they limit potential benefits from the use of AI. System implementation is permitted only after safety is confirmed and risks minimized. The obligation to prevent harm is absolute.
2.2.19. Collective Duty
The development and implementation of AI technologies essential for society is regarded as a collective obligation (fard kifayah) of the Muslim community. It is impermissible to leave society without necessary technical solutions when their creation is feasible. Responsibility for realization is distributed among all competent entities.
2.2.20. Principle of Precaution in Doubtful Matters
When doubt arises regarding the permissibility or safety of an AI system, the most cautious and safe position should be adopted. Preference is given to solutions that eliminate risks to faith, morality, and public welfare. Prudence and foresight are mandatory in all ambiguous situations.
2.2.21. Principle of Adaptation to Time and Place
Decisions and norms regarding the use of AI must be adjusted in light of changes in time, place, and circumstances. Regulatory flexibility is permitted within the limits of Shari‘ah and societal needs. Norms must not remain static if the conditions of their application have changed.
II. Higher Objectives of Shari‘ah
2.2.22. Preservation of Religion
Any use of AI that may distort the foundations of Islam, undermine faith, or impede the performance of religious rites is prohibited. All technological solutions must ensure respect for religious beliefs. Any interference in matters of faith is considered impermissible.
2.2.23. Preservation of Life
AI systems must not pose a threat to human life or health at any stage of their life cycle. Priority is given to the safety of users and others. All risks must be identified and minimized before deployment.
2.2.24. Preservation of Intellect
It is forbidden to use AI for spreading misinformation, manipulating consciousness, or creating harmful psychological dependencies. Protection of the user’s mind and consciousness is a necessary condition for system operation. Any actions undermining mental health are considered violations of Shari‘ah norms.
2.2.25. Preservation of Property
The use of AI must not result in fraud, deception, usury, or other financial abuses. All processes must be transparent and comply with the principles of fair economic activity. The protection of property rights of participants is obligatory at all stages.
2.2.26. Preservation of Lineage and Honor
AI cannot be used to encroach upon a person’s honor, reputation, privacy, or morality. Any functions that violate dignity or human rights are prohibited. Protection of family, marriage, and private life must be guaranteed by technical and organizational measures.
III. Fiqh Rules and Provisions for Various Stages of the Life Cycle
2.2.27. Honesty and Halalness of Data Sources
All data and resources used for the development and training of AI must be obtained in a lawful and honest manner. Use of information from doubtful or prohibited sources is impermissible. Verification of data origin is a mandatory procedure.
2.2.28. Prohibition of Depicting Living Beings
In robotics and visual systems, it is prohibited to create realistic images of humans and animals if this may lead to excessive veneration or social temptation. Replication of models that may violate religious or moral norms is not allowed. Responsibility for compliance lies with developers.
2.2.29. Observance of Contracts and Property Rights
It is categorically prohibited to use another’s data, technology, or intellectual property without permission. All copyright and property rights must be respected on the basis of concluded agreements. Violations of these rights are subject to immediate rectification and compensation.
2.2.30. Transparency and Absence of Gharar
All contracts and interactions with AI must be as transparent and understandable as possible to all parties. Deception, omission, unjustified risk, or misleading users are prohibited. Logging of operations and system events is mandatory.
2.2.31. Indirect Responsibility
The developer bears responsibility for harm caused by scenarios embedded in the system, even if they are realized by the end user. This responsibility is not waived by transfer of the system to third parties. All potential use cases must be evaluated at the design stage.
2.2.32. Limitation of Full Automation in Ahkam
Questions of religious law and issuance of rulings must remain the exclusive domain of humans. AI may not act as an independent judge or mufti. Any automated systems in these matters serve only as auxiliary tools.
2.2.33. Right of Choice and Appeal
The user must be granted the right of choice when interacting with AI, as well as the ability to appeal decisions made by the system. Appeal mechanisms must be transparent and accessible. Any form of imposition of decisions is a violation of user rights.
2.2.34. Consideration of Custom (Urf)
AI systems must take into account local customs and cultural specificities, provided these do not contradict Shari‘ah. Development and implementation of solutions require prior analysis of local norms. Application of general algorithms is possible only with their adaptation to specific conditions.
2.2.35. Duty to Verify Religious Data
Any religious information provided by AI must undergo mandatory verification and review by qualified specialists. Unverified or unreliable information is not allowed to be disseminated. Responsibility for data accuracy rests with both developers and system operators.
IV. Practical Guidelines for Implementation and Operation
2.2.36. Shari‘ah Audit and Oversight
AI systems must undergo regular Shari‘ah audit and oversight by authorized Islamic specialists. Audit results are documented in official records. Failure to comply with requirements results in withdrawal of operational approval.
2.2.37. Filtering of Haram Content and Functions
Effective filters must be implemented to prevent the appearance and dissemination of haram content and prohibited functions. Responsibility for filtering rests with all participants in the AI life cycle. The filtering system must be updated and improved as new risks emerge.
2.2.38. Documentation of Restrictions
All Shari‘ah restrictions must be thoroughly documented in technical and user documentation. Explanations must be made available to users and developers. Lack of transparency or concealment of these norms is considered a violation of Shari‘ah requirements.
2.3. System of Mandatory Provisions Based on the Principles of Islamic Akhlaq
2.3.1. Sabr (Patience)
Entities involved in the AI system life cycle are obliged to allocate sufficient time and resources for risk analysis and testing, refraining from deploying AI until identified deficiencies are addressed.
2.3.2. Shukr (Gratitude)
Developers of AI systems are obliged to duly acknowledge the contribution of authors of scientific works and open data, and to ensure the use of AI results for the benefit of society.
2.3.3. Tawakkul (Reliance on Allah)
In planning and evaluating AI projects, entities must combine diligence and scientific integrity with recognition of the limitations of technology, avoiding absolutization of AI capabilities.
2.3.4. Qana‘ah (Contentment)
Economic models for the use of AI must exclude excessive profit extraction that leads to unethical monopolization of data or technological dependence of users.
2.3.5. Taqwa (God-Consciousness)
Those responsible for AI systems must adhere to Islamic norms and data security standards, even in the absence of external oversight, including maintaining an internal log of actions.
2.3.6. Tawadu‘ (Humility)
AI project leaders must document model limitations and publicly report errors or malfunctions, providing for the possibility of external audit.
2.3.7. Ikhlas (Sincerity)
The goals stated in technical specifications and public communications must fully correspond to the actual intentions and functions of the AI system; hidden mechanisms are prohibited.
2.3.8. Amanah (Trustworthiness)
Entities are obliged to protect all entrusted personal and other data through up-to-date cryptographic and organizational measures, preventing unauthorized access.
2.3.9. ‘Adl (Justice)
AI algorithms are subject to mandatory verification for the absence of discrimination and bias; identified imbalances must be rectified before commercial deployment.
2.3.10. Sidq (Honesty)
AI system providers are required to supply users with truthful, complete, and comprehensible information about the capabilities, limitations, and risks of the system, including known cases of incorrect operation.
2.3.11. Ihsan (Striving for Excellence)
All stages of the AI system life cycle must be carried out in accordance with the best available industry practices and standards, aimed at the maximum public benefit.
2.3.12. Rahmah (Mercy)
Developers and operators of AI systems must provide for mechanisms to minimize harm to people and the environment and to prioritize functions that protect vulnerable groups and assist those in need.
2.3.13. Adab and Haya (Etiquette and Modesty)
AI systems intended for generating, filtering, or publishing content must block materials that contravene public morality and cultural-religious values.
2.3.14. Hikmah (Wisdom)
When making decisions regarding AI implementation, entities are required to analyze long-term social, environmental, and economic consequences with the participation of independent experts and scholars.
2.3.15. Mas’uliyyah (Accountability)
Owners and operators of AI systems bear personal and financial responsibility for the outcomes of system operation and must ensure the possibility of open external audit.
2.3.16. Ihsan fi al-‘Amal (Quality of Execution)
Each version of the AI system may be released only after documented quality control confirming compliance with stated specifications and industry standards.
2.3.17. Shura (Consultation)
Key design and regulatory decisions on AI systems are to be made collectively, with the mandatory participation of experts in Shari‘ah, engineering, ethics, and law; minutes of meetings are to be published.
3. Requirements for Design and Implementation
3.1. Design and Research Stage
3.1.1. Compliance with Shari‘ah from the Outset. At the early stages of the AI development life cycle (research, data collection, requirements analysis), the principles of Shari‘ah and akhlaq must be applied, taking into account the specifics of systems that generate various types of content (text, graphics, audio, video, software code, data), since distortion or improper generation of information may lead to the dissemination of false or haram material.
3.1.2. Multi-Faceted Risk Analysis. Developers and project managers must assess potential threats at the earliest stages, using flexible and scientifically sound methodologies (e.g., CRISP-DM for analytical projects or Data-Driven Design in combination with Shari‘ah expertise). It is important to identify risks of discrimination, privacy violations, generation of inaccurate religious content, as well as any possible moral and ethical costs.
3.1.3. Collaboration with Religious Experts. In order to conduct a “Shari‘ah filter” for projects prior to launch, it is recommended to involve theologians, ulema, and specialists in Islamic theology. Such comprehensive expertise helps avoid fundamental errors and adapt technical solutions to the norms of Islam.
3.1.4. “Design Thinking” and “Halal by Design”. The design of AI systems must be based on the concept of “halal by design”: when forming user scenarios and algorithm architecture, Islamic restrictions and requirements should be considered in order to minimize restructuring at later stages. This includes ethical interface design (UX), proper data collection (excluding unreliable or haram sources), and transparent logic.
3.2. Implementation and Operation Stage
3.2.1. Verification and Testing. Before deploying the system in a production environment, thorough testing must be conducted, including religious-ethical audits. This may include checking for the absence of discriminatory patterns, assessment of the correctness of religious multimodal content, as well as tests to exclude the exploitation of vulnerabilities (for example, for haram activities).
3.2.2. Feedback Mechanisms and Error Correction. Particular attention should be paid to rapid response to detected errors or inaccurate judgments—especially in religious and legally significant services. This involves the creation of channels for complaints, tools for rapid algorithm correction, and “hotfix” procedures to promptly address critical issues.
3.2.3. Transparency for Users. During operation, it is important to inform users about what data is collected, how it is processed, and what algorithms are used for decision-making. User agreements and interfaces must be clear, free from hidden conditions, and comply with Islamic moral norms (in particular, not encouraging prohibited actions).
3.2.4. Monitoring and Logging Systems. To ensure the safe and legitimate operation of AI, monitoring tools must be implemented (logging, anomaly tracking, behavioral analysis). The collected statistics can help detect deviations in a timely manner, for example, attempts to use the system for haram content or fraudulent operations.
3.2.5. Compliance with Regulatory Requirements. In addition to Shari‘ah norms, the operation of AI systems must take into account local data protection laws (GDPR in Europe, Federal Law 152 in Russia, etc.), as well as regulatory acts in finance, medicine, and other sectors where AI is implemented.
3.3.1. Regular Audit and Iterative Improvement. AI technologies are constantly evolving; datasets, models, and architectures change. Therefore, it is recommended to conduct systematic audits—both internal and external. This includes involving independent Shari‘ah experts and cybersecurity specialists.
3.3.2. Procedures for Algorithm Updates. Any update (for example, a new machine learning model, architecture change, or functional expansion) must be tested for compliance with Islamic norms, with particular attention to the specific requirements for generating each type of content (text, graphics, audio, video, software code, data). The process should be documented: describing how the new version of the algorithm affects the project’s risk profile, ensuring transparency for all stakeholders.
3.3.3. Emergency Intervention Mechanisms. In the event of serious violations (e.g., use of the algorithm for haram activity, improper generation of religious rulings), procedures must be in place for urgent suspension of the system or rollback to a previous stable version. This allows for minimization of potential harm to users and society.
3.3.4. Knowledge Accumulation and Best Practices. Organizations implementing AI in the Islamic context should systematize their experience and practices related to the safe updating of systems. Publishing case studies and methodological recommendations contributes to the collective growth of the industry and helps prevent the repetition of others’ mistakes.
4. Organization of Islamic Oversight
4.1. Authorized Persons for Islamic Norms in AI
4.1.1. Appointment of a Responsible Person. Each organization developing or using AI in the Islamic context shall appoint an authorized person for Islamic norms (hereinafter, “Islamic Authorized Person”) or a working group responsible for ensuring compliance with this Code.
4.1.2. Functions of the Islamic Authorized Person. He/she ensures interaction with religious councils, is responsible for collecting questions from users and developers, maintains internal documentation on matters of Islamic norms, and formulates recommendations for correcting projects and processes. The Islamic Authorized Person may participate in key meetings related to the design, launch, and updating of AI systems.
4.1.3. Qualification Requirements. It is desirable that the Islamic Authorized Person possess basic knowledge of Shari‘ah, an understanding of Islamic legal methodology, and a general understanding of AI technologies, including the principles of generating multimodal content (text, graphics, audio, video, code, data), and be able to assess their outputs’ compliance with Islamic norms.
4.1.4. Accountability and Independence. To maintain objectivity, the Islamic Authorized Person may have direct access to the organization’s leadership and should not be subject to pressure from marketing, commercial, or similar departments. His/her role is to prevent violations and help the organization follow Shari‘ah standards.
4.2. Commissions on Islamic AI Norms
4.2.1. Establishment of Commissions. Large religious or educational institutions may establish specialized commissions (or Shari‘ah councils) in the field of AI, responsible for developing methodological and legal positions and resolving complex and contentious issues.
4.2.2. Composition of Commissions. The commission should ideally include:
• Islamic theologians and muftis (Shari‘ah experts);
• IT and AI specialists;
• Legal experts (as necessary);
• Industry representatives or organizations facing practical issues in AI application.
4.2.3. Principles of Work. The work of the commissions must be based on transparency, collective discussion, and adherence to procedures of Islamic jurisprudence (usul al-fiqh). The results (fatwas, recommendations) are published openly or within corporate use to ensure public trust.
4.2.4. Functions of the Commissions. The commissions:
• Review AI application practices and identify potential Shari‘ah violations;
• Issue fatwas (legal opinions) or religious-ethical recommendations;
• Approve lists of reliable sources of religious knowledge;
• Monitor the correctness of their use in AI systems (e.g., religious chatbots);
• Advise government agencies or business structures on issues of Islamic AI norms.
4.2.5. Interaction with Organizations. If a company does not have its own Islamic Authorized Person, it may apply to such a commission for independent audit or for religious-ethical consultation on specific projects.
4.3. Register of Islamic AI Solutions
4.3.1. Purpose of the Register. To enhance transparency and trust, an open register of certified (“Shari‘ah-verified”) systems may be maintained. This register helps end users, investors, and government bodies understand which solutions comply with Islamic norms.
4.3.2. Contents of the Register. The register specifies:
• Product or service name;
• Information about the developer organization and the Islamic Authorized Person (if any);
• Brief description of functionality;
• Assessment of compliance with Shari‘ah;
• Certification status (e.g., “full compliance,” “compliant with reservations”).
4.3.3. Certification Process. Certification is usually conducted on a voluntary basis: the organization applies to the commission on Islamic AI norms or to accredited experts, who audit the solutions and issue an opinion. If the verdict is positive, the information is entered into the register. Certification may require periodic re-evaluation (re-audit) in connection with algorithm updates.
4.3.4. Advantages of Certification. Status in the register serves as an important indicator of the product’s reliability and Islamic authenticity, increasing user loyalty and creating additional guarantees for partners. Certification provides competitive advantages in markets targeting Muslim audiences.
4.3.5. Coordination with Government Agencies. In some jurisdictions, the register may be maintained under the auspices of ministries for religious affairs or authorized Islamic councils, which gives it official status. In other cases, this may be an initiative of non-governmental Islamic organizations.
5. Measures for Compliance and Development of the Code
5.1. Obligations of Participants
5.1.1. AI actors who have adopted the Code undertake to implement its prescribed norms and principles in their practice, including training staff in the basics of Islam and conducting internal audits.
5.1.2. Non-compliance with the principles may result in loss of trust from users and religious organizations.
5.2.1. Training and Professional Development Programs. Developers and project managers are advised to undergo specialized courses on Islamic norms in AI technologies. Such courses may include the basics of Shari‘ah, Islamic legal methodology, global case studies, and reviews of technical standards and regulations. Regular upskilling contributes to a deeper understanding of religious-ethical aspects.
5.2.2. Interdisciplinary Research. In order to develop Islamic AI, it is necessary to encourage research at the intersection of theology, law, sociology, and computer science. This creates a foundation for innovative solutions that take into account both technical and religious requirements. Publication in peer-reviewed journals and participation in conferences facilitate the dissemination of experience and advance the community.
5.2.3. Partnership with Educational Institutions. Universities, Islamic colleges, and research institutes can cooperate in training personnel in the field of “Islamic AI.” Joint programs, internships, and grants will help future specialists develop a holistic vision in which technology aligns with Islam.
5.2.4. Internal Seminars and Workshops. Organizations that have adopted the Code are recommended to hold regular internal educational sessions. During such sessions, Shari‘ah experts, programmers, and project managers can jointly analyze complex cases, share news about technological trends, and discuss how to apply principles of akhlaq in practice.
5.3. Experience Sharing and Cooperation
5.3.1. Formation of Professional Communities. An “Islamic AI” community (e.g., in the form of associations or clusters) can accumulate best practices, conduct joint research, and develop unified standards. Members of such a community can collaboratively create open datasets (in accordance with Shari‘ah), publish guides, and hold industry consultations.
5.3.2. Open Data and Shari‘ah Restrictions. Dissemination of open data (Open Data) contributes to scientific progress, but it is important to observe Islamic norms of privacy and ethics (not disclosing personal data, not publishing haram content). Joint initiatives will help create “halal” datasets that can be safely used to train AI models.
5.3.3. Cooperation with Islamic Financial and Charitable Organizations. As the role of Islamic banking, insurance, waqf (charitable endowments), and other religious institutions increases, interaction with them can accelerate the implementation of Islamic AI solutions. For example, joint projects in donation analytics, halal certification management, or educational platforms for the underprivileged.
5.3.4. International Conferences and Competitions. Regular forums, startup competitions, and seminars on Islamic AI expand networks and the exchange of best practices. The participation of Islamic scholars and technical experts from various countries increases diversity of opinions and helps adapt the Code to different local contexts.
5.4. Revision and Updating of the Code
5.4.1. Iterative Development. As the field of AI is rapidly changing (new machine learning methods, the emergence of quantum computing, etc.), it is necessary to constantly review the Code’s norms in light of new realities. Regular updates and involvement of experts from various fields (Shari‘ah, ethics, IT) guarantee the relevance of the document.
5.4.2. Review Procedures and Mechanisms. It is recommended to establish task forces under the commissions, which will monitor major technological trends and prepare proposals for updating the Code. It is important that this process be transparent and open to the expert community.
5.4.3. Public Consultations. Before the approval of a new edition of the Code, it is advisable to hold public hearings, inviting developers, religious experts, and representatives of civil society. This helps identify contentious issues, take different viewpoints into account, and increase the level of acceptance of the final version.
5.4.4. Support for Local Regulations. Countries and regions with a Muslim majority may issue their own standards or guidelines based on this Code. Thus, there is a continuous exchange of experience: local practice enriches the central document, and its affirmed principles set unified benchmarks.
6.1. Harmony Between Progress and Shari‘ah Norms
6.1.1. Scientific and Technological Progress in the Islamic Key. Islamic civilization has historically been renowned for its contribution to science and culture. The development of technologies, including systems generating multimodal content (texts, graphics, audio, video, software code, data), may be viewed as a continuation of the tradition of Islamic scientific progress, provided that innovations do not violate Shari‘ah norms and moral-ethical values. The introduction of new technologies is permissible only if they do not contradict the prescriptions of Islam and promote the strengthening of faith and moral standards in society.
6.1.2. Aligning Technological Progress with Islamic Values. Issues related to the adoption and regulation of AI must be resolved on the basis of the principle that technological progress is subordinate to Islamic values, not vice versa. Only this approach ensures the harmonious integration of innovative technologies with the requirements of Shari‘ah.
6.1.3. Sustainable Development and the Spiritual Dimension. The goal of the Muslim community is not merely technological leadership, but also the preservation of moral values. Therefore, any strategy related to AI must take into account multifactorial effects: economic, environmental, social, and spiritual. This ensures truly sustainable development, allowing eternal values not to be sacrificed for fleeting success.
6.2. Strengthening Trust and Improving Society
6.2.1. The Importance of Trust (Amanah). In Islam, trust between participants in social relations is of key importance. Transparent and responsible application of AI strengthens social bonds, enabling the Muslim ummah (community) to benefit from scientific achievements without fear of deception or abuse.
6.2.2. Contribution to Global Good. The dissemination of the Islamic approach to AI benefits not only Muslims, but all humanity. The Islamic tradition emphasizes the importance of justice (‘adl) and compassion (rahmah), which may become universal values for international AI initiatives.
6.2.3. Development of Civil Institutions. When AI technologies are implemented in accordance with Shari‘ah principles, they become a tool for realizing human potential: improving access to education, healthcare, and various halal services. This directly affects the overall standard of living and promotes the formation of a responsible society in which each person feels involved in the common good.
6.3. Call for Shared Responsibility
6.3.1. Interdisciplinary and Multi-Level Approach. The Islamic AI Code is addressed to a wide range of participants: theologians, scholars, businesspeople, government bodies, and users themselves. Only coordinated efforts of all stakeholders can ensure true safety and benefit of AI.
6.3.2. Collective Oversight. In Islamic law, there is a principle of oversight and mutual supervision. Applied to the sphere of AI, this entails project transparency, the possibility of external audit, and information sharing about potential risks. Thus, shared responsibility helps reduce the likelihood of systemic failures and unscrupulous use of algorithms.
6.3.3. Inter-Madhhab Cooperation. In the development and application of AI technologies in the Islamic context, it is recommended to organize consultations with representatives of various Islamic schools and directions. Such inter-madhhab dialogue helps account for all nuances of Shari‘ah interpretation and ensures broad legitimacy of decisions made.
6.3.4. Righteous Legacy for Future Generations. Technologies developed today will affect the future of the ummah. Adherence to Shari‘ah principles in the implementation of AI forms the foundation for nurturing responsible generations who will continue to improve society and science in accordance with the will of the Almighty.
6.3.5. Recognition of Global Interdependence. AI is a global phenomenon, and the Muslim community cannot consider it in isolation. This Code is intended to demonstrate readiness for cooperation and invites all interested parties (including non-Muslims) to dialogue, the exchange of best practices, and the joint pursuit of just solutions.
6.4. The Future of “Strong AI” and Transhumanism
• In the event of the development of strong AI technologies claiming consciousness or autonomy, it is necessary to uphold the inviolable principle that only a human being is endowed with a soul (nafs) and intellect (‘aql), responsible before Allah.
• Any attempts to “transfer” the human mind into a machine, “infuse” artificial consciousness into a biological body, or ascribe divine attributes to a machine are categorically contrary to ‘aqidah (monotheism).
• Transhumanist projects that blur the line between human and machine must undergo particularly rigorous Shari‘ah review, given the risk of losing moral responsibility and human nature (fitrah).
• In all circumstances, the doctrine of Tawhid remains: worship is due to Allah alone, not to created beings, no matter how perfect they may appear from a technical perspective.
This Islamic Code for the Application of Artificial Intelligence is intended to serve as a foundational document for the responsible introduction and use of AI for the benefit of Islamic society and in pursuit of the higher objectives of Shari‘ah. Its provisions may be supplemented by industry standards and local regulations detailing particular aspects of activity. However, all norms of the Code remain oriented toward the principles of justice, respect for humanity, adherence to Shari‘ah, and the pursuit of the common good.