Comment by Cyrus Hodes

Major economies should conclude bilateral and multi-lateral AI incident notification agreements—the equivalent of nuclear hotlines, requiring no new treaty body and activatable within months—so that when a regulator in one jurisdiction identifies a systemic harm, counterparts elsewhere are legally obligated to respond within their domestic capabilities rather than being left to act voluntarily and asynchronously, as happened with Grok. Unverified source (2026)