AI Vigilance is not an activity that operates solely within an Organization. Incidents, Signals, and Emerging Patterns arising from AI System Operation may require either internal handling or sharing and coordination with the External Ecosystem, including AI Providers, Regulators, Professional Communities, and other Organizations. AI Vigilance therefore includes two distinct but complementary functions: Internal Vigilance and External Vigilance.

Internal Vigilance focuses on AI Operation within an Organization. It Detects Incidents and Signals arising from Operation, collects and Assesses relevant Evidence, and Feeds Back necessary Findings into the Governance System. Continuous observation is important not only for individual Incidents, but also for Repeated Events, Deviation, Behavioral Shift, Emerging Pattern, and other changes that become visible over time.

When a detected Incident or Signal is Relevant beyond the boundaries of the Organization, connection to External Vigilance becomes necessary. External Vigilance Monitors Incidents and Signals arising across the External Ecosystem, coordinates with Regulators and Relevant Authorities, and, where appropriate, facilitates Safety Information Exchange with other Organizations. At the same time, Signals, Findings, Guidance, Safety Measures, and other information obtained externally are brought back into the Organization as Inputs to its Governance Process.

The distinction between Internal Vigilance and External Vigilance is not simply a distinction between “internal staff” and “external specialists.” The Operating Model that determines who performs Vigilance and the Functional Scope that determines how far Vigilance connects are separate design questions. Internal Vigilance may be supported by External Specialists, while External Vigilance may be performed by an Organization’s internal Team.

A key decision point for a detected Signal is: “Is this relevant beyond the organization?” If there is no Relevance beyond the Organization, Internal Handling continues. Where external impact, shared Risk, Regulatory Relevance, or other broader implications are identified, the Signal is connected to External Reporting or Coordination.

Through this structure, AI Vigilance connects Continuous Monitoring within individual Organizations with Collective Learning across Organizations. Necessary Safety Information can flow into the External Ecosystem, while knowledge generated there can return to internal Governance, forming interconnected Internal and External Vigilance Loops.