System exception and deviation reports are the most effective tools for identifying abnormal events, as they highlight deviations from expected processing and performance.
Logs (A): Provide raw data but require manual review or monitoring tools.
Operator problem reports (C): Capture issues but depend on human reporting, which may miss critical events.
Operator work schedules (D): Related to staffing, not abnormal events.

The most important regulatory consideration when using customer data to train AI algorithms is ensuring that data is collected lawfully and that explicit consent from data subjects has been obtained where required. This aligns with data protection laws such as GDPR, which emphasize lawful processing and informed consent for personal data use.

Capacity planning that omits changes to key business systems is most concerning because it can result in insufficient resources to support critical business operations, leading to performance degradation or service outages that directly impact business objectives.

Service accounts are intended for running system or application services - not for human interaction. Allowing interactive logins with service accounts increases the risk of misuse or credential compromise. Therefore, best practice is to disable interactive login for these accounts while maintaining necessary permissions for automated processes.

The greatest risk that could result from a contracted penetration tester attempting SQL injection techniques on the production system is that production data could be altered. Penetration testing activities, such as SQL injection, have the potential to manipulate or corrupt live data in the production environment. This could cause significant operational issues, data integrity problems, and business disruptions. It is essential that penetration testing is performed in a controlled environment (e.g., a staging or testing environment) to avoid any impact on production data.