The Qawqal Standard

Values should be readable.
Not hidden in an algorithm.

Qawqal is building a public, understandable values standard for the entire superapp: one baseline where faith traditions overlap, faith-aware treatment where they differ, and a clear process for difficult cases.

Shared baseline

Three lines content should not cross.

The core principles should remain simple enough for an ordinary Qawqaler to understand and specific enough to guide difficult decisions.

01 · Explicit content

No explicit sexual content.

A clear baseline that can apply across feeds, social surfaces and AI experiences, with carefully defined exceptions for legitimate contexts such as health, education and news.

02 · Modesty

Context matters.

Modesty is not identical across every faith tradition or context. Qawqal treats context and faith-aware differences as part of the problem instead of forcing one tradition's interpretation onto everyone.

03 · Faith

Disagreement is not contempt.

People can question religions, debate ideas and disagree strongly. Ridicule and contempt for sincere faith do not have to become the culture of the platform.

Where the traditions differ

Shared where possible. Distinct where necessary.

One baseline is not enough.

Muslims, Christians and Jews share important values. They also differ on important questions. Qawqal starts with a common floor without erasing legitimate differences above it.

Faith-aware. Still user-respecting.

The aim is to make the values lens understandable, give users clear settings where appropriate, and avoid surprising people with hidden assumptions.

Governance

AI screens. People govern.

A serious values platform cannot outsource every difficult moral decision to a classifier. The system needs policy ownership, human review, appeals and a visible record of what changed.

01Machine-readable policyTranslate the public Standard into rules and evaluation criteria the moderation system can actually use.
02AI screeningUse multimodal models for scale, confidence scoring and prioritization.
03Human edge casesRoute ambiguous or high-impact decisions to trained reviewers and appropriate domain advisers.
04Appeals & versionsLet difficult cases improve the policy. Important changes should create a new public version rather than disappearing silently.
The product promise

The Standard itself should earn trust.

ReadableShort enough that an ordinary Qawqaler can understand the core rules.
VersionedImportant changes should have dates, reasons and visible history.
AppealableUsers should have a path to challenge difficult decisions.
Faith-awareShared baseline without flattening real differences.
Context-awareMedical, educational, journalistic and other legitimate contexts need different treatment from entertainment.
AccountableQawqal should explain who owns the policy and how disputed cases influence it.
Why this matters strategically

The Standard can become part of the moat.

The defensibility is not the prose alone. It can come from the accumulated taxonomy, edge cases, labeled decisions, evaluation data, trust, appeals history and the operational discipline required to make a values layer actually work across an entire superapp.

Talk to us