annie
Product DesignAI Detection2026

HumanMarq

A proprietary AI-detection rating system protecting human-created comic art.

Raremarq's response to generative AI entering the collectibles market.

HumanMarq AI Launch Asset

My Role

Founding UX & Ops Designer

Timeline

February – March 2026

Tools

Figma

Team

CT

Christopher Tung

Founder

AD

Annie Ding

Founding UX & Ops Designer

JS

Jacob Singer

Co-Founder and CTO

JM

Jon

Marketing Manager

When AI Blurs the Line, Trust Erodes

As generative AI tools improved, collectors could no longer visually distinguish between:

  • Fully human-created artwork
  • AI-assisted work
  • Predominantly AI-generated pieces

This created three major risks:

Devaluation

Original comic art losing value

Trust Erosion

Buyer confidence declining

Reputation Risk

Platform credibility at stake

In a market built on provenance and scarcity, uncertainty is dangerous.

[Before vs After Placeholder]

Make Authenticity Visible

Instead of quietly moderating AI content behind the scenes, we made authenticity visible.

HumanMarq introduced:

Human-Likelihood Percentage

Quantifiable authenticity score

Transparency Spectrum

Structured clarity across categories

Zero-Tolerance Enforcement

Firm policy on AI-generated work

The goal was not just detection.

The goal was clarity.

[Detection Flow Placeholder]

Turning Detection Into Trust

The Score System

HumanMarq ranges across a spectrum:

Fully Human-Created

95-100% Human-Likelihood

Human-Assisted

70-94% Human-Likelihood

AI-Assisted

40-69% Human-Likelihood

Predominantly AI-Generated

0-39% Human-Likelihood

A spectrum builds credibility instead of binary labeling — it acknowledges nuance while maintaining clear standards.

Score Visualization

Design decisions focused on clarity and trust:

  • Clean badge system
  • Percentage displayed prominently
  • Subtext explaining "Based on automated analysis for transparency"
  • Non-accusatory language
  • Educational tone, not punitive
[Score UI Placeholder]

Iconography & States

A complete system was built to support various states:

Authenticity Badges

Visual indicators for each score tier

Confidence Indicators

Transparency about detection certainty

Dispute Pathways

Clear process for seller appeals

Ban State

For predominantly AI-generated work

Tone balance: Firm but not aggressive. Transparent but not absolute.

[Icon System Placeholder]
[Policy States Placeholder]

Product + Policy Working Together

HumanMarq wasn't just UI.

It required:

Backend Detection Signals

AI analysis infrastructure

Public-Facing Score Display

Visible trust indicators

Seller Education

Clear guidelines and expectations

Enforcement Rules

Consistent policy application

How UX Framed Policy:

Clear messaging
Appeal pathways
Trust-forward positioning

Protecting the Category We Serve

HumanMarq positioned Raremarq strategically in the market:

Differentiated Platform

Set Raremarq apart from open marketplaces

Curated Identity

Reinforced premium platform positioning

Creator Relationships

Strengthened trust with artists

Buyer Confidence

Increased trust in purchases

Category defense, not just feature launch.

Tens of thousands

of collectors

Millions

in GMV processed

Trusted

by top creators

HumanMarq Launch Graphic

What Success Looked Like

Clearer Buyer Confidence Signals

Buyers could make informed decisions with visible authenticity scores

Reduced AI-Related Disputes

Transparent scoring reduced confusion and conflict

Strong Creator Alignment

Artists appreciated the platform's commitment to protecting human work

Elevated Platform Integrity

Raremarq positioned as a trusted, quality-focused marketplace

Designing for Cultural Shifts

Trust is now a UI problem

Transparency beats silent moderation

Authenticity needs visible signals

UX can protect market value

Explore More Projects