Please enable JavaScript.
Coggle requires JavaScript to display documents.
AI and ML: Ethics, Privacy, and Cybersecurity - Coggle Diagram
AI and ML: Ethics, Privacy, and Cybersecurity
Cybersecurity Risks
Scenario 1: Employee uses AI to create convincing phishing emails
A: Use AI to trick coworkers; B: Use simulated phishing only with approval.
Choose B: Train employees without stealing information or causing harm.
Universalism: Deception would be harmful if everyone accepted it.
Fact: Generative AI can make deceptive messages sound credible.
Scenario 2: Company lets AI automatically block network traffic
A: Let AI block any suspicious traffic; B: Test, log, and require oversight for major blocks.
Choose B: Protect operations while reducing accidental service outages.
Justice/fairness: Legitimate users should not be unfairly denied access.
Fact: Automated defenses can misclassify normal activity.
Ethical Challenges
Scenario 1: AI hiring tool ranks applicants using biased past data
A: Accept rankings without checking; B: Audit results and offer human appeals.
Choose B: Give qualified applicants a fair chance.
Ethics of care: Consider applicants vulnerable to an unfair rejection.
Fact: Historical hiring patterns can reinforce discrimination.
Scenario 2: Hospital AI suggests a treatment with no explanation
A: Follow the suggestion automatically; B: Require clinician review and explain limits.
Choose B: Keep patients safe and informed.
Common-good approach: Prioritize patient trust and shared well-being.
Fact: AI predictions can be mistaken or incomplete.
Privacy and Security Challenges
Scenario 1: AI chatbot stores private student messages
A: Keep all chats for training; B: Ask consent, remove identifiers, limit retention.
Choose B: Protect privacy while still improving the tool.
Kantian ethics: Treat users as people with rights, not only data sources.
Fact: Stored chat records can reveal sensitive information.
Scenario 2: School uses facial recognition at entrances
A: Scan everyone continuously; B: Use ID cards and limited checks.
Choose B: Maintain safety without collecting unnecessary biometric data.
Rights-based ethics: Privacy should not be sacrificed without strong justification.
Fact: Biometric identifiers are difficult to replace after exposure.
Cybersecurity Advantages
Scenario 1: AI detects suspicious login attempts
A: Automatically lock every flagged account; B: Flag risk and allow human review.
Choose B: Stop attacks while reducing unfair account lockouts.
Utilitarianism: Compare prevented attacks with harm to innocent users.
Fact: Detection systems can produce false positives.
Scenario 2: AI finds vulnerabilities in company software
A: Publish exploitable details immediately; B: Report privately and allow time to fix.
Choose B: Help defenders repair problems before attackers exploit them.
Virtue ethics: Responsible disclosure shows care and good judgment.
Fact: Vulnerability reports can help patch weaknesses.