SECAI Core · phase 15 of 15
Rollback
To restore a system to the last known good state.
The Explain card
- Plain English
- A rollback restores a system to the last known good state, undoing a change that broke something or introduced a risk.
- Example
- A newly deployed chatbot version starts leaking fragments of its system prompt after a prompt update. The team rolls back to the previous version within minutes while they investigate.
- Why it matters
- Every AI change is a bet. Fast, tested rollback turns a bad model or a harmful prompt tweak from a crisis into a brief blip, but only if previous versions are kept and the rollback is rehearsed.
- Hook
- The undo button for production: test it before you need it.
Where it sits in the deck
Phase 15: MLOps, Continuous Delivery, and Operational Resilience
Close the loop — operational practices, drift management, and delivery discipline that keep deployed AI systems healthy, current, and continuously improving.
