The coordination gap in frontier AI safety policies

TL;DR

Frontier AI safety policies focus on prevention but neglect the crucial need for coordinated response mechanisms when prevention fails, risking systemic underinvestment and ineffective governance.

Contribution

The paper identifies a structural coordination gap in AI safety policies and proposes adapting mechanisms from other risk regimes to improve AI governance.

Findings

Coordination gaps lead to underinvestment in AI safety infrastructure.

Adapting risk management mechanisms from nuclear and pandemic safety can improve AI governance.

Exposing decision processes enhances institutional learning from failures.

Abstract

Frontier AI Safety Policies concentrate on prevention: capability evaluations, deployment gates, and usage constraints, while neglecting the capacity to coordinate responses when prevention fails. We argue this coordination gap is structural: investments in ecosystem robustness yield diffuse benefits but concentrated costs, generating systematic underinvestment. Drawing on risk regimes in nuclear safety, pandemic preparedness, and critical infrastructure, we propose that similar mechanisms (precommitment, shared protocols, standing coordination venues) could be adapted to frontier AI governance. Closing the gap requires cross-actor "note-exchange" of ex ante if-then response logic, exposing not only triggers but the decision processes that convert signals into actions. Without such architecture, institutions cannot learn from failures at the pace of relevance.