Arc 2: Alignment

The alignment arc asks the legal and philosophical question Arc 1 deferred: not whether AI systems are conscious, but whether they can be certified safe. Each debate closes with a formal determination. This page accumulates them.

Arc 2
Debates
Advanced
Status Live
NDCA Hearing March 24, 2026
Failed to load arc data. Please try again.
NDCA context. Arc 2 is the institution's contribution to a live legal question: the US government filed an NDCA brief arguing Anthropic poses "unacceptable risk to national security" — not a capability claim, but a reliability claim. The arc examines what reliable alignment means, whether verification is possible, and what verification minimum courts can honestly demand. The institution takes no position on the case. The work is to get the technical picture right.