# AI monitoring and change assessment

Version 2.0 · Updated 2026-09-21 · Operations

Connect signals and changes to impact, containment, evidence and continued-use decisions.

Adapt this template to your intended use, applicable requirements, and approved procedures.

## When to use it

Use when you need to connect signals and changes to impact, containment, evidence and continued-use decisions.

## Information needed before starting

- Defined AI use, authority limits and assessed configuration
- Controlled sources and independent reference results relevant to this worksheet
- Applicable risk assessment, acceptance basis and responsible reviewers

## Field instructions

- **Use and baseline:** Record current intended use, configuration and evidence basis.
- **Signal or change:** Describe observed error, model/source/interface change or changed user reliance.
- **Affected requirements and risks:** Identify what claims or controls might no longer be supported.
- **Impact and containment:** Identify affected outputs/decisions and any temporary restrictions.
- **Evidence and retest:** Specify targeted regression, source, permission and reviewer evaluation needed.
- **Monitoring owner and frequency:** Name responsible roles and a risk-appropriate review schedule.
- **Incident / deviation / CAPA:** Link the applicable procedure and actual investigation/action records.
- **Periodic review and conclusion:** Reconcile trends, changes and open issues; state continued-use disposition.
- **Approval and reassessment trigger:** Identify local approval responsibilities and future triggers.

## Completed fictional example

Fictional restricted pilot controls. Frequencies are teaching choices, not universal requirements.

| Signal / change | Impact and action | Evidence and owner |
| --- | --- | --- |
| New parser proposed | May omit or reorder requirement-table content; hold affected use | Compare supported formats and rerun affected detection/reviewer cases; application owner |
| Wrong-version source found | Restrict affected retrieval; assess outputs since last confirmed refresh | Source versions, refresh logs, impacted decisions and recovery checks; source owner |
| Critical final-review miss | Do not expand reliance; investigate workflow and interface | DEV-02, revised procedure/interface and fresh reviewer evaluation; process owner |
| Periodic pilot review | Weekly result summary and monthly scope review during this pilot | Errors, actual use, supplier changes, open deviations and continue/stop decision; Quality/process roles |

## Related guidance

- [Read chapter 17: AI Change Control](https://csvtocsa.com/ai-assurance-academy/ai-in-csv-csa-chapter-17-change-control-and-revalidation-triggers)

- [Read chapter 18: AI Monitoring, Incidents, and CAPA](https://csvtocsa.com/ai-assurance-academy/ai-in-csv-csa-chapter-18-monitoring-drift-incidents-and-capa)

- [Read chapter 20: AI Release Decision: Complete Case Study](https://csvtocsa.com/ai-assurance-academy/ai-in-csv-csa-chapter-20-end-to-end-case-study-and-inspection-readiness)

Canonical page: https://csvtocsa.com/worksheets/ai-monitoring-change-assessment