Page 59 - Ai V2.0 Flipbook C7
P. 59
Transparent, Explainable
People should know the purpose of data collection and agree to it. This helps others trust the
findings and check them if needed. For example, if you are collecting survey answers, people
should know the purpose and agree to it.
Fair and Unbiased
Using data from only certain people or places can create unfair or one-sided results. Data
should not be used to harm or discriminate. For example, the automation in the process of job
selection should not reject applications based on the discrimination against certain races. It
should help everyone, not hurt a specific group.
Privacy and Data Protection
Data must be protected, and individuals' privacy should be respected. It must not be shared
without their approval. For example, data collection should not reveal anyone's identity when
surveying a mall's busiest hours.
Accountable
Accountability means taking responsibility for your actions, especially when things go wrong.
In statistical data, data collection plays a key role. The person who collects the data, the data
analyst, and the decision maker are all equally responsible if something goes wrong. For
example, if the government uses test score data to rank schools, then it should be clear who
would be accountable if any school is ranked low due to incomplete or biased data.
Safe, Secure, and Sustainable Transparent, Fair and Privacy and
Explainable Unbiased Data Protection
Statistical data related to safety, such as
crime rates, road accidents, or workplace
issues, can be biased if the data is not
collected fairly, key information is missing, Accountable Safe, Secure,
and
or it is incorrectly interpreted. Sustainable
21 st Century #Media Literacy
Skills
Video Session
Watch the video of "Ethics of AI: Challenges and Governance" at the given link:
https://youtu.be/VqFqWIqOB1g?si=WNFTIqSTjeKanaaT OR scan the QR code and
answer the following question:
Can you point out other ethical issues related to AI which are not discussed
in the chapter?
AI and Ethical Considerations 57

