Module 1: Why AI Ethics Matter

In 2018, Amazon scrapped an AI recruiting tool that had been in development for four years. The system, trained on a decade of hiring data, had learned a simple pattern: Amazon historically hired mostly men for technical roles. So the AI concluded that being male was a positive signal. It systematically downgraded resumes containing words like "women's" (as in "women's chess club captain") and penalised graduates from all-women's colleges.

Nobody programmed it to be sexist. Nobody intended the outcome. The data was biased, the system learned the bias, and it took years to notice. Amazon's recruiting team is sophisticated — if they can build a sexist AI accidentally, anyone can.

This is why AI ethics isn't a philosophical luxury. It's a practical necessity.

The Scale Problem

When a single human recruiter is biased, they affect dozens of candidates. When an AI system is biased, it affects millions. That's the fundamental ethics equation: AI multiplies both the benefits and the harms of any decision at unprecedented scale.

COMPAS, a criminal sentencing algorithm used across US courts, was found by ProPublica to falsely label Black defendants as high-risk at nearly twice the rate of white defendants. This system influenced sentencing decisions for over a million people before the bias was publicly documented.

Scale turns small errors into systemic injustice. That's why we can't treat AI ethics as an afterthought.

The "It's Just a Tool" Fallacy

You'll hear this constantly: "AI is just a tool. It's neutral. Ethics depend on how people use it."

This is wrong, and dangerously so. Here's why:

Tools have values embedded in their design. A hammer doesn't, but a recommendation algorithm that optimises for engagement over accuracy has a value baked in: attention matters more than truth. That's not neutral.

AI makes autonomous decisions. A spreadsheet doesn't deny your mortgage application. An AI system can and does — often without explanation. When a tool makes decisions, it needs ethical scrutiny.

Opacity obscures responsibility. When an AI denies someone a loan, who's responsible? The developer? The company? The data scientists? The training data? This ambiguity is a feature of AI systems, not a bug — and it's an ethical problem.

What AI Ethics Actually Covers

AI ethics isn't one thing. It's a cluster of interconnected issues:

• Bias and fairness — Who gets harmed when AI reflects societal prejudices?


• Privacy and surveillance — What happens when AI can track, identify, and predict everyone?


• Transparency — Can people understand and challenge AI decisions that affect them?


• Safety — How do we prevent AI systems from causing harm?


• Accountability — When AI causes damage, who's responsible?


• Labour and economics — What do we owe people whose livelihoods AI disrupts?


• Environmental impact — What's the carbon cost of training AI systems?


• Autonomy — When should AI make decisions, and when should humans?

This course covers all of these. Not as abstract philosophy, but as practical problems with real consequences.

Why You Should Care (Even If You're Not Building AI)

You don't need to build AI to be affected by AI ethics. You're already affected:

• Your social media feed is shaped by algorithms optimising for engagement


• Your credit score may incorporate AI-driven assessments


• Your job application may be screened by AI before a human sees it


• Your insurance premiums may be influenced by AI risk models

Understanding AI ethics makes you a better citizen, a better professional, and a harder person to exploit. Ignorance isn't protection — it's vulnerability.

---