AI Safety: Not Optional, Not Later

2026-09-09

AI Summary

论文提出一种结合模型级监督和系统级控制的安全设计架构，以解决多层AI安全故障问题。

Abstract

Incidents show that AI safety failures often arise across multiple layers. We present a safety-by-design assurance architecture combining model-level supervision, such as Scientist AI, with system-level controls over scaffolds and harnesses, independent verification, monitoring, and evidence infrastructure, supported by governance for accountability and evidence interoperability.

Problem

• AI Safety
• Safety Failures
• System-level Controls

Innovation

• safety-by-design
• model-level supervision
• system-level controls
• governance for accountability