AI for Islamic Social Finance and Development teaches how artificial intelligence can, and often should not, improve the delivery of zakat, waqf, sadaqah, qard hasan, Islamic microfinance and humanitarian assistance to the people who need them most. It is the sibling of our blockchain programme and shares its discipline: begin with the human need and the Maqasid, locate the delivery problem, weigh honest alternatives, and only then ask whether AI genuinely adds value. Every module carries an AI-Honesty Test, because knowing when not to use AI matters as much as designing with it. Across fourteen modules the course runs at three levels, from understanding the concepts, to applying them to real workflows, to designing and governing AI in an institution. It teaches the durable ideas, models, data, hallucination, agents, autonomy, oversight and governance, that outlast any single tool, and it excludes machine-learning engineering, hype and speculation, keeping its focus on technology that serves development.

Islam measures a work not by its sophistication but by whom it serves. This course places AI in its proper station, beneath the Maqasid and in service of the poor, the displaced and the excluded, and it treats beneficiary data as an amanah rather than raw material. Its recurring question, whether AI truly helps or merely dazzles, is an act of sidq, a refusal to overstate. Two lines are held firmly throughout: an AI system may retrieve and summarise rulings from authentic sources but must never issue a fatwa or invent one, and the fiduciary, Shariah and moral amanah for every decision remains with the human institution and its scholars, never the model. A learner leaves able to build, and equally able to say honestly when the simpler path is the better one.