Check Point says it cut vulnerability triage from hours to minutes with Daybreak.
Can a model be trusted to decide what gets fixed first?
The short answer from OpenAI is yes, if customers keep approval for important actions.
Check Point says a small analyst team once tracked every public flaw by hand, so it now continuously monitors customer networks for gaps and lets each customer define intent for the secure state it wants.
Daybreak pairs OpenAI models with Codex Security to identify threats, generate patches and verify remediation across code and systems, per its product page.
Check Point warns attackers now wield similar tools, so watch false positives and real patch times before calling hours-to-minutes a win.




