TOP SECRET // CLASSIFIED // INTEL 2027

HUMANITY'S
ENDGAME

A Global Strategy Game

An asymmetric card game of AI governance, espionage, and catastrophe. Five hidden roles. Six rounds. One permanent deadline: January 1, 2030.

5
HIDDEN ROLES
6
ROUNDS
2030
DEADLINE

// INTEL FILE

THE FIVE ROLES

🧠
THE LAB
OpenBrain CEO

Reach Capability 10 before rivals.

ACCELERATE: +2 CAP, +1 Risk
⚖️
THE REGULATOR
AI Safety Commissioner

Keep global Risk below 60 through Round 6.

AUDIT: Force a hand reveal. Risk –1
📢
THE WHISTLEBLOWER
Ex-OpenBrain Researcher

Expose 3 secrets to trigger a global Slowdown.

LEAK: Reveal a Secret Card. Draw 1
🕵️
THE ADVERSARY
DeepCent Director

Steal 2 Lab Capability cards undetected.

INTERCEPT: Copy any card played this round
🚀
THE ACCELERATIONIST
Billionaire Techno-Optimist

Push global CAP above 8 before Round 4.

FUND: Give any player +2 CAP

// BRIEFING

HOW IT WORKS

01
CHOOSE YOUR ROLE

Pick one of five asymmetric roles — each with a unique win condition, secret objective, and special ability. No two games play the same.

02
PLAY YOUR HAND

Each round, simultaneously select a Decision Card from your hand. Cards shift the Global Risk Meter and each player's Capability score.

03
WATCH THE CLOCK

The Global Risk Meter ticks up every round. If it hits 100, humanity loses. Counter tokens, role abilities, and Intel Votes can shift momentum.

04
HIT YOUR WIN CONDITION

The Lab needs CAP 10. The Regulator needs Risk below 60. The Whistleblower needs 3 exposed secrets. First to their condition wins.

// DATA PRODUCT

GOVERNANCE ALIGNMENT INDEX

Aggregate win-rate data from millions of games becomes a real-time index of how the public thinks AI governance is going. Academic and corporate API access coming in Season 2.

GAI SCORE
Unlocks after first 10,000 games