Educerie · IB Diploma · Digital Society

Topic 1 — Concepts

The eight concepts are the lenses you look through. They are not a list to memorise and recite — they are the tools that turn "this technology is bad" into an argument that earns marks.

1.1 Change, expression and identity

Change — digital systems alter how people live, work and relate, and the rate of change is itself part of the story. Two distinctions matter. Change can be incremental (each phone a little better) or disruptive (streaming destroying the video rental industry). And change is uneven: the same technology arrives at different times and on different terms for different groups, which is what the digital divide describes — differences in access, in the quality of that access, and in the skills to use it.

The strong analytical move is to ask who benefits from the pace of change. Rapid deployment favours those who build systems; slower deployment favours those who must adapt to them.

Expression — digital systems widen who can publish and what forms expression can take. Anyone can broadcast. The tension is between freedom of expression and harm: the same absence of gatekeepers that lets a dissident publish lets disinformation spread, and any moderation regime that stops the second constrains the first. There is no configuration that maximises both, and saying so explicitly is a mature answer.

Identity — digital systems shape how people present and understand themselves. Online identity can be curated, plural, anonymous or pseudonymous. Anonymity protects the vulnerable and shields the abusive, in the same design decision. Consider also imposed identity: profiling systems assign you a category — a credit band, a risk score, an advertising segment — that you did not choose, may not know about, and often cannot contest.

1.2 Power, space and systems

Power — who decides, who benefits, and who is accountable. This is the most useful concept in the course and the most under-used by candidates.

Ask three questions of any system: Who owns it? Who is affected by it? Who can change it? The answers usually differ, and the gap between the second and third is where the argument is. Platform users are affected by ranking algorithms they cannot see, do not control and cannot appeal. Concentration of power matters too: a handful of firms control search, mobile operating systems and cloud infrastructure, which makes them effectively regulators of activity they do not formally govern.

Space — digital systems change the meaning of place. Work, education and healthcare detach from location; a person can be present somewhere they are not. But space also reappears in a hard form: data has a physical location, and data sovereignty — which country's law governs data stored on which soil — is a live political question. Physical geography persists in cables, data centres and the mineral extraction that supplies devices.

Systems — digital technologies are interconnected and interdependent, which produces effects nobody designed. A change in one part propagates; a failure in one part cascades. When a single cloud provider fails, unrelated services fall over together. Systems thinking is also what makes unintended consequences predictable as a category even when the specific consequence is not.

1.3 Values and ethics

Values are what individuals and societies hold to be important; ethics is reasoning about what is right. Digital dilemmas are almost always conflicts between values that are each defensible, which is why they persist.

The recurring conflicts:

• Privacy versus security. Surveillance that prevents harm also enables control. The question is never whether to trade off, but how much, under what oversight, and who decides.
• Openness versus safety. Open data enables research and accountability; it also enables re-identification and misuse.
• Innovation versus precaution. Deploying early captures benefits sooner and discovers harms in production, on the public.
• Efficiency versus fairness. An algorithm optimising a defensible objective can produce an indefensible distribution.
• Individual versus collective. Contact tracing benefits the population by exposing individuals.

Three ideas worth using by name:

Informed consent is the standard, and it is rarely met. Consent obtained through a forty-page agreement nobody reads, with no genuine alternative to accepting, is consent in form only. Saying why it fails — length, complexity, absence of alternative, bundling of unrelated permissions — is stronger than asserting that it does.

Algorithmic bias arises chiefly from data, not from malice. A model trained on historical decisions reproduces the pattern of those decisions; if past hiring favoured one group, a model learning from it will too, and will do so at scale and with the appearance of objectivity. The examinable point is that this is a systemic problem requiring auditing and diverse data, not a bug to be patched.

Accountability is the hardest of the three. When an autonomous vehicle causes harm, responsibility is distributed across the manufacturer, the software developer, the training data, the regulator, the owner and possibly the person inside. Distributed responsibility tends to become no responsibility — which is why liability frameworks matter more than good intentions.

How to build an answer with these

A reliable four-step structure for any extended question:

1. Name the concept the question turns on — power, identity, values.
2. Ground it in a real, named example, with a date.
3. Give at least two stakeholder perspectives, and explain why their interests diverge.
4. Conclude with a judgement that depends on something specific — who bears the risk, whether consent was meaningful, whether an intervention can actually be enforced.

What actually loses marks in this topic

1. Writing about the technology instead of about its social effects.
2. Generic examples. Name the company, the system, the year.
3. Listing concepts without applying them.
4. Presenting one perspective as though it were the only reasonable one.
5. Treating algorithmic bias as deliberate rather than as a property of training data.
6. Asserting that consent was not informed without explaining why.
7. Concluding with "there are pros and cons" instead of a judgement.
8. In Paper 2, not quoting or referring to the sources provided.

Educerie · written from the published IB syllabus structure for Digital Society Topic 1, first assessment 2024. Original text. Last reviewed 5 September 2026.