Detect and Mitigate Ethical Risks 

Detect and Mitigate Ethical Risks Course Introduction

CEET Specialization Introduction

Course Welcome & Success Tips

The Importance of Managing Risks

Risk Management Process

Risk Identification

Risk Analysis

Risk Mitigation

Types of Ethical Risk

Distributions

Central Tendency

Variance and Standard Deviation

Skewness and Kurtosis

Correlation

Probability

Machine Learning Outcomes

Cost Functions

Reliability

Goodhart's Law

Overview

Risk Management Frameworks

Classification Metrics

Analyzing Ethical Risks

The Importance of Managing Privacy Risks

Private Data

First-Party vs. Third-Party Data

Secondary Use of Data

Combined Data Sources

Identify Personally Identifiable Information (PII)

Model Personas

Track Customer Data

Meet Compliance Requirements

Intent and Consent

Minimize Private Data Sharing

Give the User Choices

Minimize Private Data Collection

Reinforce Trust

Anonymization and Pseudonymization

Homomorphic Encryption

Zero-Knowledge Protocols

Parity Introduction

Incorporate Privacy Risk Management in the Lifecycle

Overview

Data Protection Policies

Privacy Legislation Sources

Managing Privacy Risks

The Importance of Managing Accountability Risks

Use of Third-Party Components

Automation Bias

Extrajudicial Judgment

Lack of Guiding Principles

Recognize Black Box Algorithms

Assess the Organization's Governance Structure

Document and Distribute Company Policies

Document Design Processes

Document Auditing Processes

Responsibility Assignment Matrix (RAM/RACI)

Pilot Testing

Collaboration with Data Sharing Partners

Algorithmic Impact Assessment (AIA)

Data Visualization

Dashboard Reporting

Incorporate Accountability Risk Management in the Lifecycle

Overview

Managing Accountability Risks Quiz

The Importance of Managing Transparency and Explainability Risks

Black Box Systems

Self-Learning Models

Third-Party Integration

Intellectual Property Rights

Shadow Banning

Explainable AI

Identify Algorithmic Decisions

Deconstruct Specific Decisions

Explain How Systems Work

Help Users Seeking Explanations

Keep Humans in the Loop

Ensure Proper Data Disclosure

Be Upfront About Training Data Inadequacies

SHAP and Alibi

ELI5, LIME, and What-If

Overview

Incorporate Transparency and Explainability Risk Management in the Lifecycle

Managing Transparency and Explainability Risks Quiz

The Importance of Managing Fairness and Non-Discrimination Risks

Implicit Bias

Sampling Bias

Reinforcement Bias

Temporal Bias

Overfitting to Training Data

Edge Cases and Outliers

Analytical Techniques

Analyze Models in Different Environments

Persona Modeling

Inclusive Design and Foreseeability

STEEPV Analysis

Perform User Testing

Gather Input from External Stakeholders

Bias and Safety Bounties

AI Fairness 360

Radioactive Data Tracing

Incorporate Fairness and Non-Discrimination Risk Management in the Lifecycle

Overview

Pattern Matching vs. Bias

AI Fairness 360 Demo

Managing Fairness and Non-Discrimination Risks

The Importance of Managing Safety and Security Risks

Abnormal System Behavior

Adversarial Machine Learning

Bad Actors

Groupthink and Biases

Cyber Attacks

Quantitative Risk Analysis

Evaluate Training Data and Models

Threat Intelligence

Threat Modeling

Penetration Testing

Forensic Analysis

Ensure Critical AI Systems Follow Rigorous Standards

Establish Baseline System Behavior

Designate Rapid Response Teams

Protect the Security of Data in Storage

Protect the Security of Data in Transit

Threat and Risk Libraries

Threat Modeling and Analysis Tools

Attack Simulation Tools

Vulnerability Scoring Tools

Security Information and Event Management (SIEM)

Incorporate Safety and Security Risk Management in the Lifecycle

Overview

Managing Safety and Security Risks Quiz