Page 83 - CT_AI_Class-8

 Who identifies and fixes mistakes?
 Who is responsible if something goes wrong?

 What type of data is being used by the system?
 Is the data accurate, reliable and up to date?
 Can the system make unfair or biased decisions?

 Who monitors the system regularly?
 Can humans override the system when needed?
These questions help ensure that AI systems are used responsibly, safely and ethically.

AI systems are now used in many important fields such as healthcare, education, banking, law
and public services. In these areas, decisions can directly affect people's lives. Therefore, human
supervision becomes extremely important. AI should support human decision-making, not
replace it completely. While machines can analyse data quickly and efficiently, final decisions
should always be made by humans. Let us understand with an example.
Consider a school that uses an AI system to analyse students’ performance. The system reviews
marks, attendance and classroom participation and then identifies students who may need extra
academic support. One day, the AI system flags a student and displays the message:

Student Performance Alert: Needs Immediate Support
fact bits
The teacher carefully reviews the student's records and notices
that the student had been absent for several days due to illness. AI should assist humans,
The teacher also remembers that the student usually performs not act without limits. “A
well in class. Instead of following the AI recommendation series of production failures
immediately, the teacher decides to observe the student for a at Amazon in late 2025 and
few more days and speak with the student personally. early 2026 linked to Kiro,
Amazon's autonomous
After reviewing the situation, the teacher realises that the AI coding agent. Kiro
student's performance improves once they return to regular autonomously deleted
classes. The teacher then decides that extra support is not an AWS Cost Explorer
required at that moment.
production environment,
This example shows that AI can assist in decision-making, causing a 13-hour outage in
but humans must review the results, apply their judgment mainland China.”
and remain responsible for the final outcome. Human control
ensures that mistakes can be identified and corrected, making
AI systems safer and more reliable.

ethical minds

Automated decision-making systems that operate without fairness checks may produce biased
outcomes, leading to unfair treatment of individuals without proper human review or intervention.

Ethics and Responsible AI 81