Please enable JavaScript.
Coggle requires JavaScript to display documents.
Privacy & Security Challenges, Cybersecurity Advantages, Cybersecurity…
Privacy & Security Challenges
Data Collection
Surveillance in Smart Homes
Health Monitoring Apps
Consent & Transparency
Hidden data use by fitness trackers
Unclear data retention policies
Scenario: An AI health app shares patient data with third parties.
Options:
Share anonymized data
Share data with full identity
Choice: Share anonymized data only
Why: Based on Virtue ethics – promotes trust and responsibility
Fact: HIPAA restricts identifiable health data sharing
Scenario: A smart speaker collects and stores voice data without user knowledge.
Options:
Leave it unregulated
Require transparent opt-in consent
Choice: Require transparent consent
Why: Supports Deontological ethics – duty to respect autonomy
Fact: GDPR requires informed consent for data collection
Cybersecurity Advantages
Threat Detection
Real time malware recognition
AI in phising email filter
Incident Response
Faster breach reaction
Automated threat mitigation
Scenario: AI blocks a legitimate user assuming it’s a threat.
Options:
Prioritize security over access
Balance access with improved verification
Choice: Improve verification
Why: Utilitarianism – best outcome for the most people
Fact: False positives can reduce system trust
Scenario: AI detects harmful code and deletes files without confirmation.
Options:
Immediate deletion
User-confirmed response
Choice: User-confirmed response
Why: Care ethics – considers user experience and context
Fact: Human-in-the-loop is a cybersecurity best practice
Cybersecurity Risks
Adversarial Attacks
Image manipulation in facial recognition
Poisoned training data
Bias & Misidentification
AI wrongly flags people or behavior
Over-policing in minority communities
Scenario: A biased AI system flags more users from minority groups.
Options:
Keep using the system
Retrain with diverse data
Choice: Retrain with diverse data
Why: Justice theory – fairness and equity
Fact: Studies show facial recognition misidentifies Black faces 5x more often
Scenario: AI recommends locking user accounts due to suspicious activity caused by VPN.
Options:
Immediate lock
Review before action
Choice: Review before action
Why: Social contract theory – systems should protect users' rights
Fact: VPN usage is not always malicious
Ethical Challenges
Autonomy vs Automation
Self-driving cars deciding accident outcomes
AI job replacement
MIsinformation
Deepfakes and false news
Manipulated media influencing elections
Scenario: Self-driving AI must choose between harming passengers or pedestrians in a crash.
Options:
Protect passengers
Minimize total harm
Choice: Minimize harm
Why: Utilitarianism – choose the lesser of two evils
Fact: MIT’s Moral Machine study explored this dilemma
Scenario: AI generates fake news that goes viral.
Options:
Allow full AI freedom
Add editorial review
Choice: Editorial review required
Why: Deontological ethics – duty to truth
Fact: Deepfakes can sway public opinion during elections