Purpose before technology
We begin with service to Islam and Muslim communities, then choose the technology. Novelty alone is never a sufficient reason to build.
A working charter
AI assists. People remain responsible.
We use AI where it can provide useful, testable help. Sound knowledge, integrity, and human responsibility remain foundations that no model replaces.
Knowledge stays with its experts
Religious and educational material is built and reviewed with qualified people. An automated model does not become an independent authority.
People make consequential decisions
AI may suggest, organize, or summarize. Decisions that affect people or sensitive content remain under accountable human judgment.
Privacy by necessity
We collect only what a task requires, restrict access, and do not use sensitive information to train models without explicit authority and a clear justification.
Limits are stated plainly
We do not present probability as certainty or claim accuracy and impact that testing has not established. People should know when they are engaging with AI-generated output.
Test before scaling
We examine quality, bias, error, and foreseeable misuse, beginning in a limited setting before expanding a system.
Security from the start
Permissions, safeguards, and accountability are designed at the beginning—not after data has accumulated or use has grown.
Review is continuous
Models, risks, and needs change. We therefore revisit system performance and safeguards and correct what proves inadequate.
These are working commitments applied when we choose, design, test, and operate a project.
Share an idea that can serve Islam and Muslims