Ethics and Technology

Beyond "Does It Work?"

Every topic in this course has asked some version of "how do we build this correctly?" — a correct algorithm, a working program, a secure system. Ethics asks a different, equally important question: should we build this, and how do we build it responsibly? Technology is not neutral in its effects — it touches real people, and understanding computing means understanding that responsibility too.

Privacy: Who Knows What About You

Modern software collects data constantly — location, contacts, browsing habits, voice recordings, photos. Collecting data is not automatically wrong; the ethical questions are what is collected, why, whether people were clearly told, and whether they meaningfully agreed. An app quietly selling location history to advertisers, buried in a wall of unread terms and conditions, is different from a fitness app openly using location to plot your running route — same data, very different ethics, hinging entirely on informed consent.

This connects directly to online safety and encryption: the tools protecting your data only matter if organisations holding it also choose to respect it.

Fairness: When Systems Treat People Differently

Artificial intelligence showed how a system trained on unrepresentative data can perform worse for underrepresented groups — a facial-recognition system misidentifying some people more than others, a hiring tool learning patterns from a biased history of past decisions. The core insight generalises well beyond AI: any system reflects the assumptions and data of the people who built it, and those assumptions can unintentionally disadvantage people who were not considered during design. Recognising this is the first step toward building — and demanding — fairer systems.

Access: Not Everyone Starts From the Same Place

The digital divide describes real, persisting gaps in who has reliable devices, fast internet, and the skills to use them confidently. When schools, job applications, or government services move online — often for good reasons of efficiency — anyone on the wrong side of that divide can be quietly locked out. Ethical technology design asks who might be excluded, not just who is easiest to serve.

Accessibility is the closely related design question: does software work for people with visual, hearing, motor, or cognitive differences? Screen-reader-friendly layouts, captions on video, alternative ways to provide input beyond a mouse — these are not optional extras bolted on afterward, but design decisions made (or neglected) from the start, shaping who can actually use what gets built.

Responsibility: When Things Go Wrong

Software has bugs; systems get breached; malware exploits vulnerabilities. The ethical question is not whether failures ever happen — they will — but how organisations respond. Discovering that user passwords were exposed and promptly, honestly informing affected people (so they can change passwords, watch for fraud) is a fundamentally different choice than quietly hiding the breach and hoping nobody notices. Transparency after failure is a core professional responsibility, not an optional courtesy.

Genuine Trade-Offs, Not Easy Answers

Some questions in technology ethics resist a single "correct" answer, and pretending otherwise is itself dishonest. Encryption protects ordinary people's private messages — and also shields criminals' communications from investigators. Powerful AI tools can help enormously and can also displace jobs or spread convincing misinformation. Data that improves a recommendation system also builds a detailed profile of a person's habits. These are genuine value trade-offs — reasonable, well-informed people can weigh privacy against security, or innovation against risk, differently. Recognising a trade-off as a trade-off, rather than pretending one side has no cost, is itself the ethical skill.

A Habit, Not a Chapter

The practical takeaway is a habit worth carrying into every future project, not a fixed list of rules: before building something, ask who benefits, who could be harmed, who was and wasn't consulted, and what happens when it inevitably fails. These questions do not have code-based answers — but asking them consistently is what separates technically skilled computing from responsible computing.

Try It Yourself

Pick one app you use daily. List: what data does it likely collect, was that clearly explained to you, who might be disadvantaged by how it works, and what would responsible failure-handling look like if it were breached tomorrow?

Worked Example — Working Through a Genuine Design Dilemma

Ethical reasoning becomes concrete when applied to an actual decision, not just abstract principles. Imagine a school is deciding whether to install facial-recognition cameras at the gates, sold as a way to speed up entry and improve safety. Walk through the questions this lesson recommends, one at a time:

• Who benefits? Staff save time on manual sign-in; the school can claim a safety improvement.
• Who could be harmed? As covered in artificial intelligence, facial-recognition systems can perform less accurately for some groups than others if trained on unrepresentative data — meaning misidentification is not evenly distributed as a risk.
• Was informed consent sought? Students and families may not have been clearly told exactly what data is stored, for how long, or who else might access it.
• What happens if it fails? A data breach here would expose not just names but biometric data — arguably more sensitive and harder to "reset" than a leaked password, since a face cannot be changed the way a compromised password can.

Notice that working through these questions does not automatically produce a single "correct" answer — different schools, with different community values and different risk tolerances, might reasonably reach different conclusions. What this process does guarantee is that the decision is made with the real trade-offs visible, rather than adopted purely because the technology exists and promises convenience. That visibility, not a guaranteed right answer, is what this lesson means by "a habit, not a chapter."

Key Words

• Informed consent — people knowingly and meaningfully agreeing to how their data is used
• Bias — systematically unequal treatment or performance across groups, often from unrepresentative data
• Digital divide — unequal access to devices, connectivity, and digital skills
• Accessibility — designing technology usable by people with a wide range of abilities
• Trade-off — a genuine tension between competing values with no single "correct" resolution