← Back to AIAF home

In development · Reader interest

AIAF Trust Card: the concept

Front and back of the proposed matte-black AIAF Trust Card, with green accents and example founding number 0001
Front and back design concept. Number 0001 is illustrative; no card or number is reserved.

A physical connection to your AIAF identity. We are exploring a card that could open an optional member profile through a QR code or NFC tap.

Founding 1,000

The proposed founding edition would feature individual numbers from 0001–1000, a founding-member badge and an optional personalised avatar. These are ideas under development, not confirmed launch benefits. A founding number would mark early membership; it would not certify someone's trustworthiness.

Identity is not permission

A name, avatar, membership number or verified identity can help answer who is this? It does not answer what may they—or their AI agent—do? Those are separate trust questions.

A future agent-facing Trust Card or trust record would need to answer five evidence questions:

  1. Who acted? Identify the agent, who controls it and what has actually been verified.
  2. What was authorized? Preserve the specific action the person approved.
  3. What limits applied? Show spending, merchant, data-access, time, confirmation and revocation rules.
  4. What happened? Record what the agent attempted or completed, linked to the authority under which it acted.
  5. Did behaviour remain trustworthy? Watch for drift or unusual patterns that should narrow the limits, seek confirmation or stop the action.

These questions reflect the direction of emerging work on user intent and agentic payments. Mastercard’s September framework connects identity, intent, controls, execution and intelligence; its first agent-detection score is planned for US testing and is not proof of authorization. The W3C Digital Credentials API remains a Working Draft focused mainly on credentials about people, not delegated agent authority. EMVCo and FIDO are also developing relevant specifications. These are signals to watch—not proof that AIAF has a certified trust product. Read AIAF’s evidence briefing.

Start small. Prove usefulness.

Before deciding on a prototype, we are defining the problem and watching the standards. The first physical-card test, if approved, would assess only tapping, QR fallback and control over profile visibility. It would not authorise an AI agent, approve a payment or certify that a person or machine is trustworthy.

Any later agent-permission prototype would be separate and clearly labelled experimental. It would need testable permission limits, expiry and revocation—not merely a badge or attractive profile.

Interested in a card?

Message AIAF on Facebook with “Trust Card interest” and the one activity you would want the card to help with. This expresses interest only: it does not reserve a card or number. No payment or identity documents are needed.

Messages are handled through Facebook under its privacy settings. There is no automatic waitlist or guaranteed launch notification yet. Follow AIAF for public updates.

Launch date, price, availability and final benefits are not set. We will publish details before accepting reservations.