PreferencePassport — Portable Taste, Context & Consent Layer for Consumer AI Integrations

The Problem

Consumer personalization is becoming more visible, more conversational, and more cross-platform.

That changes the job to be done.

Historically, apps could keep their recommendation logic inside the product:

• infer taste silently,
• optimize rankings internally,
• and expose only the end result.

That model weakens when users increasingly expect to:

• steer recommendations in natural language,
• bring their taste into external assistants,
• understand why something was recommended,
• and decide what context is shared with which AI surface.

Most apps are not built for that.

They have recommendation models and account linking, but they do not have a robust layer for:

• portable preference graphs,
• context-specific recommendation permissions,
• explanation trails for AI-generated recommendations,
• or creator-safe controls around what downstream assistants can access.

The opportunity is to build the consent-aware preference operating layer for consumer AI integrations.