CLASSIFIED — TEAM EYES ONLY
National Security Council & Executive Branch
You are the most powerful government on Earth, but power doesn't mean control. You regulate AI, control chip exports, deploy military and intelligence assets, and manage alliances. But you answer to voters, face corporate lobbying from the very labs you're trying to regulate, and operate through a slow, bureaucratic system in an exponentially fast-moving domain.
Maintain AI strategic leadership over China. Keep domestic stability above crisis levels. Preserve global influence and alliance networks. Ensure deployed AI systems remain aligned with human values.
Your technological advantage over China. Driven by export controls, compute investment, lab partnerships. If this drops below 40, China has reached parity.
Public trust, economic health, social cohesion. Affected by job displacement, AI incidents, political polarization. Below 30 = constitutional crisis territory.
Alliance strength, institutional control (UN, NATO, G7), soft power. Unilateral actions erode this. Multilateral ones build it.
Your trust that deployed AI is safe. Based on safety testing, lab transparency, interpretability results. This is the metric no one talks about publicly but everyone should be watching.
Information known only to your team
Strategic competition with espionage on both sides. You have the upper hand in compute but they're closing the gap. A hot conflict would be catastrophic for everyone.
They're your national champion in the AI race, but they resist regulation, lobby against your export controls, and may be hiding safety concerns. You need them more than they need you.
They share your safety concerns and cooperate with your agencies. But they're a company, not a patriotic institution — they'll prioritize survival over your agenda if pushed.
NATO ally, shared values. But the EU AI Act creates compliance nightmares for your labs, and Europe wants a seat at the table you'd rather keep small.