What is 'explainability' in the context of an AI-powered SOC tool, and why is it specifically important for audit trail purposes in regulated industries?

• A.Explainability is only relevant for customer-facing AI decisions under GDPR; internal SOC tools have no explainability requirement
• B.Explainability means the AI's source code is published as open-source so regulators can inspect it
• C.Explainability refers to the ability to articulate why the AI made a specific decision (e.g., 'alert flagged because user accessed 500 files in 10 minutes, 4 standard deviations above their 30-day baseline') - in regulated industries this is required because incident response actions must be justified by documented rationale, not 'the AI said so'
• D.An explainable AI model is one that achieves over 95% accuracy on a validation set, making its behavior predictable