Page 35 - CMA Journal (Mar-Apr 2025)

Focus Section

Islamic Ethics and Their Relevance to AI to earn a livelihood and warns against economic
systems that marginalize the vulnerable.
Islamic ethics are derived primarily from the Qur'an, the
Sunnah (traditions of the Prophet Muhammad P.B.U.H), 4) Deepfakes and Misinformation: AI-generated
Ijma' (consensus), and Qiyas (analogical reasoning). content such as deepfakes can be used to spread
These sources emphasize values such as justice (adl), misinformation, defame individuals, or incite
accountability (mas'ooliyyah), privacy (sitr), public violence -- activities clearly condemned in Islam
interest (maslahah), and the prohibition of harm (la darar under the injunctions against fitnah (sedition) and
wa la dirar). buhtan (slander).
These principles are directly applicable to the ethical Core Principles for Islamic AI Ethics
challenges posed by AI, such as biasness in data and
1) Human Dignity and Autonomy: Islamic teach-
algorithms, lack of transparency in decision-making,
threats to privacy, and the replacement of human ings uphold the dignity of every human being
(Qur’an 17:70). AI applications must not reduce
judgment with machine learning systems.
human beings to mere data points or deny them
Let us briefly align key Islamic ethical principles with agency. For example, AI systems should aid human
contemporary concerns in artificial intelligence. decision-making rather than override it without
consent or explanation.
• Justice (Adl): Algorithmic fairness, data neutrality,
and the avoidance of biasness in both – algorithms 2) Accountability and Transparency: The concept
and data. of hisab (accountability) is central to Islamic morality.

• Accountability (Mas'ooliyyah): Traceability of AI Every decision made by or through AI must be audit-
able and traceable to human oversight. AI systems
decisions to humans.
must be designed with transparency mechanisms to
• Privacy (Sitr and Hurmah): Protection of person- prevent abuse and enable redress.
al data and dignity.
3) Justice and Equity: Allah commands believers to
• Public Welfare (Maslahah): Ensuring AI applica- stand firmly for justice (Qur’an 4:135). AI systems
tions contribute to societal good and do not cause must be scrutinized to ensure they do not reinforce
societal harm. systemic injustice. Equity audits, impact assess-
ments, and inclusion in datasets are vital.
• Prohibition of Harm (La Darar): Avoidance of AI
misuse in surveillance, warfare, and manipulation. 4) Public Welfare (Maslahah): In Islamic jurispru-
dence, laws and policies must promote the welfare
The Ethical Challenges of AI in the Context
of the ummah (community). AI should be harnessed
of Pakistan to improve healthcare, education, access to justice,
and public services -- especially for underserved
1) Data Privacy and Surveillance: In Pakistan,
communities.
concerns around unauthorized data collection, facial
recognition without consent, and lack of transparen- 5) Precautionary Principle (Sadd al-Dhara’i):
cy in surveillance initiatives are rising. While AI-driv- Islamic ethics advise closing doors to harm before
en surveillance may serve security interests, it must they materialize. This principle supports the careful
be weighed against the Islamic principle of hurmat and proactive assessment of AI risks. For example,
al-insan (dignity of the human being) and the right autonomous weapons or invasive surveillance AI
to privacy. should be approached with extreme caution.
2) Bias and Discrimination: AI algorithms trained How to Observe Islamic AI Ethics in Pakistan
on unrepresentative data can perpetuate biases,
leading to discriminatory outcomes in hiring, Given the relevance of Islamic ethics to Pakistan’s cultural
lending, or law enforcement. In a society like and religious ethos, it is both feasible and beneficial to
Pakistan, where ethnic, linguistic, and sectarian embed these principles into the country’s AI ecosystem.
diversity exists, biased AI systems can exacerbate This can be done through the following channels.
inequalities, violating the Islamic principle of adl
1) National AI Ethics Framework Anchored in
(justice).
Islamic Values: Pakistan should develop an AI
3) Automation and Employment: As AI systems ethics framework that explicitly references Islamic
automate labour-intensive processes, there is fear of values alongside international standards (such as the
mass unemployment, especially among low-skilled UN Resolution, EU AI Act, OECD, UNESCO, and ISO
workers. Islam places a strong emphasis on the right 42001). This framework should:

ICMA’s Chartered Management Accountant, Mar-Apr 2025 33