Definition
AI tools used in cybersecurity that show their reasoning step-by-step — explaining why they flagged something as a threat — rather than just outputting a verdict with no explanation. Security vendors are building this into automated alert-triage tools specifically so human analysts can trust and audit the AI's decisions.
Why it matters
As companies hand more security decisions to AI, leaders need to know those decisions can be checked and justified after the fact — an unexplainable AI verdict is hard to defend to auditors, regulators, or a board after an incident.