AIGP EXAM WITH 100% CORRECT
ANSWERS
What aspects does the "Economic Context" dimension analyze? - Answer- Sector
Environment: Financial, healthcare, education, etc.
Business Characteristics:
• Actual business function
• AI system model type
• Criticality to operations
Deployment Factors:
• How it was deployed
• Impact of deployment
• Scale of the system
Maturity Level: Newer systems = less testing; Mature systems = more effective
What does the "Data and Input" dimension focus on? - Answer- Data Types: What kind
of data was used in the model
Expert Input: Human knowledge codified into rules
Key Characteristics:
• Data collection methods (machine vs. human)
• Data structure and format
• Collection methodology
What does the "AI Model" dimension cover? - Answer- Technical Type: What kind of AI
model is it?
Model Construction: How the model is built
Model Usage: How the model is used
Obligations for GPAI models include: - Answer- - Maintaining technical documentation
- Transparency: making information available to downstream providers who integrate
the GPAI model into their AI systems
- Complying with EU copyright law
- Providing summaries of training data
- Authorized representative: must be established in the EU and appointed by written
mandate (similar to GDPR EU representative)
,Focus: The technical architecture and implementation
What does the "Tasks and Output" dimension examine? - Answer- Tasks: What the AI
system performs
Outputs: Results produced by the system
Actions: What happens as a result of outputs
Key Elements:
• Individual tasks
• Combined task systems
• Evaluation methods for performance assessment
How would you apply the OECD Framework to evaluate an AI system? - Answer- Step-
by-step approach:
1. People & Planet: Identify all stakeholders and impacts
2. Economic Context: Determine sector, criticality, and maturity
3. Data & Input: Assess data sources and expert knowledge
4. AI Model: Understand technical architecture
5. Tasks & Output: Define functions and evaluation methods
Goal: Comprehensive risk assessment and classification
The OECD AI Framework uses dimensions to help organizations and .-
Answer- "The OECD AI Framework uses [5] dimensions to help organizations [classify
AI systems] and examine risks."
The 5 dimensions: People & Planet → Economic Context → Data & Input → AI Model
→ Tasks & Output
Why is the "maturity" aspect important in the Economic Context dimension? - Answer-
Newer systems: Less testing over time, potentially less reliable
Mature systems: More data exposure, typically more effective
Risk implication: Maturity level affects risk assessment and deployment decisions
Evaluation factor: Important for determining appropriate oversight and monitoring
What are the 6 modern drivers of AI and Data Science? - Answer- 1. Cloud Computing
2. Mobile Technology & Social Media
3. Internet of Things (IoT)
4. Privacy-Enhancing Technologies (PETs)
5. Blockchain
6. Computer Vision, AR/VR, and Metaverse
,How does Cloud Computing drive AI and Data Science development? - Answer- -
Enables scalable computing resources for AI model training
- Reduces infrastructure costs for organizations
- Provides accessible AI services and platforms
- Supports collaborative development across teams
- Facilitates rapid deployment of AI solutions
How do Mobile Technology and Social Media contribute to AI advancement? - Answer-
- Data Explosion: Generate massive amounts of user data
- Rich Information Sources: Provide AI models with diverse learning material
- Real-time Data: Enable continuous model improvement
- User Behavior Insights: Offer patterns for AI to learn from
- Global Reach: Create worldwide datasets for training
What role does IoT play in AI and Data Science? - Answer- - Massive Data Generation:
IoT devices create continuous streams of data
- Real-world Data: Provides practical, operational information
- Sensor Networks: Enable comprehensive environmental monitoring
- Edge Computing: Supports distributed AI processing
- Model Development: Offers rich datasets for AI training
What are Privacy-Enhancing Technologies (PETs) and why are they important for AI? -
Answer- Technical solutions addressing personal data and privacy concerns
Enable responsible AI and data science growth
- Benefit: Allow data use while protecting privacy
- Impact: Ensure continued AI development within regulatory frameworks
- Three Main Categories: Cryptographic, Data Minimization, Identity & Access
Management
, What are the three main categories of Privacy-Enhancing Technologies (PETs)? -
Answer- 1. Cryptographic Technologies: Protect data through encryption methods
2. Data Minimization Technologies: Reduce privacy risks while maintaining utility
3. Identity and Access Management: Control access and protect identity information
What are the three key Data Minimization Technologies and their functions? - Answer-
1. Differential Privacy
- Adds mathematical noise to datasets
- Prevents individual identification
- Preserves statistical utility
2. Federated Learning
- Trains ML models across decentralized data sources
- Avoids centralizing sensitive data
3. Synthetic Data Generation
- Creates artificial datasets
- Maintains statistical properties without real personal information
How does Blockchain contribute to AI and Data Science? - Answer- Secure financial
transactions interface
Enhances data privacy and security in certain contexts
Limitation: Not universally applicable to all data privacy and AI challenges
Security Benefit: Provides tamper-resistant data storage
Trust Mechanism: Enables secure data sharing between parties
What is Computer Vision and how does it impact AI development? - Answer- Enables
machines to understand the world through images and videos
Categories of GPAI - Answer- Chapter V of the Act lays down a legal framework for two
types of general-purpose AI:
2. General-purpose AI models
3. General-purpose AI models with systemic risk (definition is based on computing
power and substantial compliance requirements)
- Assessing model performance
- Assessing and mitigating systemic risks
ANSWERS
What aspects does the "Economic Context" dimension analyze? - Answer- Sector
Environment: Financial, healthcare, education, etc.
Business Characteristics:
• Actual business function
• AI system model type
• Criticality to operations
Deployment Factors:
• How it was deployed
• Impact of deployment
• Scale of the system
Maturity Level: Newer systems = less testing; Mature systems = more effective
What does the "Data and Input" dimension focus on? - Answer- Data Types: What kind
of data was used in the model
Expert Input: Human knowledge codified into rules
Key Characteristics:
• Data collection methods (machine vs. human)
• Data structure and format
• Collection methodology
What does the "AI Model" dimension cover? - Answer- Technical Type: What kind of AI
model is it?
Model Construction: How the model is built
Model Usage: How the model is used
Obligations for GPAI models include: - Answer- - Maintaining technical documentation
- Transparency: making information available to downstream providers who integrate
the GPAI model into their AI systems
- Complying with EU copyright law
- Providing summaries of training data
- Authorized representative: must be established in the EU and appointed by written
mandate (similar to GDPR EU representative)
,Focus: The technical architecture and implementation
What does the "Tasks and Output" dimension examine? - Answer- Tasks: What the AI
system performs
Outputs: Results produced by the system
Actions: What happens as a result of outputs
Key Elements:
• Individual tasks
• Combined task systems
• Evaluation methods for performance assessment
How would you apply the OECD Framework to evaluate an AI system? - Answer- Step-
by-step approach:
1. People & Planet: Identify all stakeholders and impacts
2. Economic Context: Determine sector, criticality, and maturity
3. Data & Input: Assess data sources and expert knowledge
4. AI Model: Understand technical architecture
5. Tasks & Output: Define functions and evaluation methods
Goal: Comprehensive risk assessment and classification
The OECD AI Framework uses dimensions to help organizations and .-
Answer- "The OECD AI Framework uses [5] dimensions to help organizations [classify
AI systems] and examine risks."
The 5 dimensions: People & Planet → Economic Context → Data & Input → AI Model
→ Tasks & Output
Why is the "maturity" aspect important in the Economic Context dimension? - Answer-
Newer systems: Less testing over time, potentially less reliable
Mature systems: More data exposure, typically more effective
Risk implication: Maturity level affects risk assessment and deployment decisions
Evaluation factor: Important for determining appropriate oversight and monitoring
What are the 6 modern drivers of AI and Data Science? - Answer- 1. Cloud Computing
2. Mobile Technology & Social Media
3. Internet of Things (IoT)
4. Privacy-Enhancing Technologies (PETs)
5. Blockchain
6. Computer Vision, AR/VR, and Metaverse
,How does Cloud Computing drive AI and Data Science development? - Answer- -
Enables scalable computing resources for AI model training
- Reduces infrastructure costs for organizations
- Provides accessible AI services and platforms
- Supports collaborative development across teams
- Facilitates rapid deployment of AI solutions
How do Mobile Technology and Social Media contribute to AI advancement? - Answer-
- Data Explosion: Generate massive amounts of user data
- Rich Information Sources: Provide AI models with diverse learning material
- Real-time Data: Enable continuous model improvement
- User Behavior Insights: Offer patterns for AI to learn from
- Global Reach: Create worldwide datasets for training
What role does IoT play in AI and Data Science? - Answer- - Massive Data Generation:
IoT devices create continuous streams of data
- Real-world Data: Provides practical, operational information
- Sensor Networks: Enable comprehensive environmental monitoring
- Edge Computing: Supports distributed AI processing
- Model Development: Offers rich datasets for AI training
What are Privacy-Enhancing Technologies (PETs) and why are they important for AI? -
Answer- Technical solutions addressing personal data and privacy concerns
Enable responsible AI and data science growth
- Benefit: Allow data use while protecting privacy
- Impact: Ensure continued AI development within regulatory frameworks
- Three Main Categories: Cryptographic, Data Minimization, Identity & Access
Management
, What are the three main categories of Privacy-Enhancing Technologies (PETs)? -
Answer- 1. Cryptographic Technologies: Protect data through encryption methods
2. Data Minimization Technologies: Reduce privacy risks while maintaining utility
3. Identity and Access Management: Control access and protect identity information
What are the three key Data Minimization Technologies and their functions? - Answer-
1. Differential Privacy
- Adds mathematical noise to datasets
- Prevents individual identification
- Preserves statistical utility
2. Federated Learning
- Trains ML models across decentralized data sources
- Avoids centralizing sensitive data
3. Synthetic Data Generation
- Creates artificial datasets
- Maintains statistical properties without real personal information
How does Blockchain contribute to AI and Data Science? - Answer- Secure financial
transactions interface
Enhances data privacy and security in certain contexts
Limitation: Not universally applicable to all data privacy and AI challenges
Security Benefit: Provides tamper-resistant data storage
Trust Mechanism: Enables secure data sharing between parties
What is Computer Vision and how does it impact AI development? - Answer- Enables
machines to understand the world through images and videos
Categories of GPAI - Answer- Chapter V of the Act lays down a legal framework for two
types of general-purpose AI:
2. General-purpose AI models
3. General-purpose AI models with systemic risk (definition is based on computing
power and substantial compliance requirements)
- Assessing model performance
- Assessing and mitigating systemic risks
