Consequences of uses of computing – A-level Computer Science exam questions with mark scheme

1Core3 marks

Laws often struggle to keep up with developments in computing.

State three challenges that legislators face when making laws about the use of computers and the Internet. [3]

Any one from: technology changes faster than laws can be written and passed, so laws quickly become out of date; the Internet crosses national borders, so data / servers / offenders may be in another country (jurisdiction); offenders can be anonymous / hard to trace; laws are hard to enforce or evidence is hard to gather; must balance privacy against security / crime prevention; must balance freedom of speech against harmful content; legislators may lack technical understanding(1)
A second, different challenge from the list above(1)
A third, different challenge from the list above(1)
Answer: Any three: technology changes faster than law; cross-border jurisdiction; anonymity of offenders; hard to enforce; balancing privacy and security / free speech; lack of technical understanding
Examiner tip: Each challenge must be a different idea; "the Internet is global" and "servers are abroad" are the same jurisdiction point.

2Standard4 marks

A free fitness app records each user's location, heart rate and daily routine. The company that makes the app analyses this data and sells it to advertisers.

Explain two risks to individuals of the app collecting and analysing this personal information. [4]

First risk identified, e.g. loss of privacy / users are monitored / a detailed profile of the user is built up(1)
Explanation, e.g. location and routine show where the user lives and when they are away from home, so they could be targeted by stalkers or burglars(1)
Second, different risk identified, e.g. data could be stolen in a data breach / data sold or shared without users understanding / health data used against the user(1)
Explanation, e.g. stolen data could be used for identity fraud / insurers or employers could use heart-rate data to refuse cover or jobs / users receive manipulative targeted advertising(1)
Answer: Any two explained risks, e.g. loss of privacy revealing home location and routine; data breach or sale of health data used against the user
Examiner tip: Each risk needs a consequence for the user to get the second mark; "it is a privacy risk" on its own scores one mark.

3Standard4 marks

A software engineer is writing the algorithm that chooses which posts appear in each user's feed on a social media app used by millions of people.

(a) Explain how the algorithm could embed the moral or cultural values of the people who design it. [2]

(b) Explain why the large scale of the app increases the responsibility of the engineer. [2]

(a) Designers decide what the algorithm rewards / what counts as a 'good' post (e.g. maximising engagement or time on the app, or what content is ranked down or removed)(1)
(a) So their own judgements / priorities / cultural norms are built into which content users see (e.g. sensational content is promoted; views from some cultures are hidden), even though users may not share those values(1)
(b) Any decision in the code is applied to millions of users at once / automatically(1)
(b) So a small error or bias can cause harm on a very large scale (e.g. spreading misinformation, harming mental health, influencing elections); allow: scale can also do great good, so the engineer must consider the effects carefully(1)
Answer: (a) designers choose what is promoted, so their values/priorities shape what users see (b) decisions affect millions at once, so errors or bias cause harm on a large scale
Examiner tip: In (a) link the designer's choice to its effect on what users see; just saying "the algorithm is biased" is not enough.

4Hard9 marks

Police forces are trialling live facial recognition cameras in public places to identify suspects.

Discuss the ethical, legal and social issues raised by this use of computing. [9]

Benefit: faster identification of wanted people; may deter crime and improve public safety(1)
Privacy: people are monitored and identified without consent; chilling effect on freedom of assembly(1)
Accuracy and bias: higher error rates for some groups can cause wrongful stops or arrests(1)
Legal / data protection: face images are biometric personal data; processing must be lawful, necessary and proportionate, stored securely and not kept longer than needed; laws may lag behind the technology(1)
Transparency and accountability: who sees the data, who checks the system, how can people challenge errors?(1)
Balanced conclusion: acceptable only with safeguards, oversight and clear limits(1)

Level	Marks	What it looks like
Level 3	7–9	Discusses ethical, legal and social issues, each developed with reasons or examples, considering both benefits and risks; uses accurate technical terms; a coherent, logically structured discussion reaching a justified conclusion.
Level 2	4–6	Several relevant issues explained, but one of ethical, legal or social is missing or underdeveloped, or the discussion is mostly one-sided; any conclusion is only partly supported.
Level 1	1–3	A few relevant points stated with little explanation; no real discussion or conclusion.

Answer: Discussion with a judgement; see indicative content
Examiner tip: Cover all three areas the question names (ethical, legal, social) and finish with a justified judgement; a list of points without evaluation stays in Level 1–2.

5Hard6 marks

Increasing automation and artificial intelligence are replacing some jobs in warehouses and offices.

Discuss the effects of this on workers and on society. [6]

Some routine jobs lost, causing unemployment or need to retrain(1)
New jobs created (maintenance, development, data roles) but needing different skills(1)
Higher productivity and lower costs, possible cheaper goods(1)
Widening inequality between skilled and unskilled workers / regional impact(1)
Safer work as machines do dangerous tasks; shorter hours possible(1)
Weighing up, e.g. overall benefit depends on retraining, education and support for displaced workers (credit a supported conclusion, but one is not required for full marks)(1)
Indicative content is not exhaustive: credit other valid effects, e.g. AI used to monitor and score workers' productivity (privacy/stress), bias embedded in automated decisions about staff, deskilling of remaining roles, responsibility of developers/employers for the impact of the systems they deploy(1)

Level	Marks	What it looks like
Level 3	5–6	Effects on both workers and society are explained with examples, considering both benefits and drawbacks; a coherent, logically structured line of reasoning. A conclusion is not required, but any conclusion given is supported by the discussion.
Level 2	3–4	Several relevant effects explained with partly developed reasoning; the answer may focus mainly on workers or mainly on society, or be mostly one-sided (benefits only or drawbacks only).
Level 1	1–2	A few relevant effects stated with little or no explanation. (0 marks: nothing worthy of credit.)

Answer: A balanced discussion of effects on workers (job losses in routine roles, need to retrain, new technical roles needing different skills, safer work) and on society (higher productivity and cheaper goods, risk of widening inequality and regional decline); a supported conclusion is credited but not required.
Examiner tip: Cover both workers and society and give benefits as well as drawbacks; a brief supported conclusion helps but breadth and explanation earn the marks.

6Standard4 marks

A bank uses an algorithm trained on past loan decisions to decide whether to approve loans.

Explain two ways in which the algorithm could produce unfair decisions. [4]

First way identified, e.g. training data reflects past human bias or is unrepresentative of some groups(1)
Explanation: the algorithm repeats or amplifies that discrimination, so some groups are refused more often(1)
Second, different way identified, e.g. inputs such as postcode act as a proxy for protected characteristics (e.g. ethnicity) / the designers chose what to optimise (e.g. profit) so their values are built in / data about an applicant is wrong or out of date(1)
Explanation, e.g. applicants from some areas are refused regardless of their own circumstances / people who would repay are refused because the model favours low-risk profit / the applicant is judged on inaccurate information(1)
Answer: Any two explained, e.g. biased training data repeats past discrimination; proxy variables such as postcode disadvantage some groups
Examiner tip: Each way needs a cause and a consequence for two marks.

Still losing marks on this topic?

Book a free introductory session and we'll find exactly which step is costing you.

Book a free intro