AI Alliance Drafts Confidential Incident Reporting Rules With Public Deadlines

The Linux Foundation's Open Secure AI Alliance proposed the Shared AI Findings Exchange (SAFE), requiring members to report AI security incidents confidentially and then publicly within set deadlines. The framework defines reportable AI incidents, notification timelines, and mandates sharing near misses and control-failure analyses to improve AI security practices.