Scribd
Upload
17 views2 pages

Tutorial 10 - AI & Governance PDF

The document discusses the Model AI Governance Framework from Singapore, emphasizing its algorithm-agnostic nature, which applies to all AI systems regardless of the specific algorithms used. It highlights the importance of internal governance structures for defining roles and responsibilities in AI management, as well as the potential downsides of human involvement in AI systems, such as increased bias and reduced efficiency. Additionally, it includes exercises related to risk classification and considerations for bias in AI design and implementation.

Uploaded by

analindatoh.at
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
17 views2 pages

Tutorial 10 - AI & Governance PDF

The document discusses the Model AI Governance Framework from Singapore, emphasizing its algorithm-agnostic nature, which applies to all AI systems regardless of the specific algorithms used. It highlights the importance of internal governance structures for defining roles and responsibilities in AI management, as well as the potential downsides of human involvement in AI systems, such as increased bias and reduced efficiency. Additionally, it includes exercises related to risk classification and considerations for bias in AI design and implementation.

Uploaded by

analindatoh.at
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Official (Open)

GRC

1. Algorithm-agnostic means that the framework applies to all AI-based


systems , regardless the type of AI algorithm used to implement the
Tutorial 10 system . this means that the method of learning , processing and working
should not affect the implementation of the framework.

AI Ethics & Governance 2


Exercises agnostic = dc which one

1. The Model AI Governance Framework from Singapore is said to be algorithm-


agnostic. Explain.
It is because It does not focus on specific AI methodology. It applies to the design, application and use of
AI in general

2. Explain why Internal Governance Structures is important, within the Model AI


Governance Framework?
Internal Governance structure is important bcause clear roles and responsibilities need to be defined to monitor and
manage the implementation, use and maintenance of AI systems. INterna; gpvernance structure help to ensure that such
roles and responsibilities are defined , assigned and enforcedwithin the organization
3. In involving humans in the functioning of an AI system in order to improve
trustworthiness, certain downsides can be expected. Name TWO of them.
Two downsides are : 2. AI systems need to formar and display relevant information to the
1. INcreased risk of human biaseness human in order to facilitate him/her to make a decision. This requires
2. reduced efficiency and scalability more elaborate design
4. Based on the Risk Impact Matrix, classify the following into the green, yellow,
orange and red zone. At the same time, indicate the human involvement model
recommended.

Sno Use Case Zone Model


a. An AI system helping a doctor
by suggesting diagnosis of orange human in the loop

condition
b. Shopping product green human out of the loop
recommendation
c. Posting recommendation yellow human over the loop
system in a social media green human out of the loop
application
d. Chat bot system in a bank human out of the loop
green
customer service
yellow human over the loop
5. Consideration of biasness in an AI system should begin at
a) Design stage
b) Implementation stage
c) Deployment stage
d) Training stage

6. Human centricity includes the following except


a) Transparency
b) Trustworthiness
c) Explainability
d) Communicability

GRC Page 1
Official (Open)
GRC

7. A company implements an AI system that seems to provide promising results.


However, it is unable to determine why. This is an issue of:
a) Explainability
b) Trustworthiness
c) Transparency
d) Reliability

2.

GRC Page 2

You might also like