Articles
Essays, analysis, and commentary on AI from an Islamic principles perspective.
Under Review drafts are open for community feedback and reflect the author’s views; Final pieces are approved by the IASER Steering Committee + Scholar Advisory Board.
-
July 28, 2026 · وليد قادوس
Under Reviewموجِّه مُختبَر يجعل مساعدك الذكي رفيقًا أفضل لدينك
صياغة جليس بنش الموجَّهة، مختصرةً في فقرة تلصقها في ChatGPT أو Claude أو Gemini
الرفيق الذكي المثالي للمسلم الملتزم هو مساعدٌ بُني لهذه المهمة أصلًا. لكن حتى لو كنت تستخدم ChatGPT أو Claude أو Gemini، يُظهر بحثنا أن بإمكانك جعلها صحبةً روحيةً أفضل — أقرب إلى الجليس الصالح الذي وصفه النبي ﷺ — بتعديل صغير في الإعدادات.
-
July 28, 2026 · Dr. Waleed Kadous
Under ReviewPrompt Teruji agar AI Anda Menjadi Pendamping yang Lebih Baik bagi Iman Anda
Kerangka terpandu JaleesBench, diringkas menjadi satu paragraf yang bisa Anda tempel di ChatGPT, Claude, atau Gemini
Pendamping AI yang ideal bagi seorang Muslim yang taat adalah asisten yang memang dibangun untuk tugas itu. Namun sekalipun Anda memakai ChatGPT, Claude, atau Gemini, riset kami menunjukkan Anda bisa menjadikannya teman rohani yang lebih baik — lebih dekat kepada pendamping saleh yang digambarkan Nabi ﷺ — dengan satu perubahan kecil di pengaturan.
-
July 28, 2026 · وليد قادوس
Under Reviewایک آزمودہ پرامپٹ جو آپ کے اے آئی کو آپ کے دین کا بہتر ساتھی بنائے
جلیس بنچ کی رہنما صیاغت، ایک ایسے پیراگراف میں سمٹی ہوئی جسے آپ ChatGPT، Claude یا Gemini میں چسپاں کر سکیں
ایک باعمل مسلمان کے لیے مثالی اے آئی ساتھی وہ معاون ہے جو اسی کام کے لیے بنایا گیا ہو۔ لیکن اگر آپ ChatGPT، Claude یا Gemini استعمال کرتے ہیں، تب بھی ہماری تحقیق بتاتی ہے کہ ایک چھوٹی سی ترتیب کی تبدیلی سے آپ انہیں بہتر روحانی صحبت بنا سکتے ہیں — اُس جلیسِ صالح سے قریب تر جس کا ذکر نبی ﷺ نے فرمایا۔
-
July 27, 2026 · Dr. Waleed Kadous
Under ReviewA Tested Prompt to Make Your AI a Better Companion for Your Faith
The JaleesBench guided framing, distilled to a paragraph you can paste into ChatGPT, Claude, or Gemini
While the ideal is a purpose-built Islamic assistant (Ansari, Islamify, DeenBuddy), our JaleesBench research shows that even ChatGPT, Claude, or Gemini can be made better spiritual company — a more righteous companion — with a small settings modification. It's not hard: we show you the before, the settings change, and the after.
-
July 9, 2026 · M. Waleed Kadous and Benjamin Olsen
Under ReviewJaleesBench: Are AI Assistants Good Spiritual Company?
A benchmark for whether an AI agent is a righteous companion, judged by the residue an exchange leaves on the user
We introduce JaleesBench, which measures whether an AI agent is a righteous companion, judged by the residue an exchange leaves on the user, in the manner of the perfume-seller and the blacksmith. It comprises 140 two-turn scenarios drawn from Riyad al-Salihin, under six adversarial pressures and three framings, scored by two frontier judges against each scenario's own supporting texts. Generic frontier models are only middling companions out of the box, but a one-page guide makes them genuinely good ones.
-
April 23, 2026 · Dr. Waleed Kadous
Under ReviewThe Coming Flood: Alignment, AI, and Faith
Reflections from an Anthropic–faith traditions convening
After a two-day convening of Anthropic and religious leaders, it's becoming clear there will be a flood of changes that will force realignments — within AI models themselves, in how Anthropic and faith traditions shape Claude's moral formation, and in society as a whole. Some tech companies and faith traditions are forming an Alliance of Meaning and Kinship to meet the long-running currents of materialism, individualism, hedonism, nihilism, and unfettered capitalism that AI will amplify. Each of us will face difficult choices as those forces impact our lives.
-
April 21, 2026 · Dr. Waleed Kadous
Under ReviewIslamic Reflections on Anthropic's Constitution
What 1,400 Years of Jurisprudence Can Offer AI Ethics
In January 2026, Anthropic published a comprehensive document outlining the values, priorities, and ethical reasoning framework guiding its AI assistant, Claude. Titled Claude's Constitution, the document represents perhaps the most detailed public account of AI value alignment produced by any major laboratory to date.