Ethics & Ownership
Writing code that works is only half of being a computing professional, the other half is judging what that code should be allowed to do, who owns it, and what happens when the two collide. This chapter covers professional ethical bodies, the real-world impact of ethical and unethical decisions, why copyright legislation and software licensing exist, and the social, economic and environmental impact of artificial intelligence. This chapter covers Cambridge 9618 syllabus section 7.1.
Cambridge groups this entire chapter under a single syllabus subsection, 7.1, but it packs in five genuinely different ideas: why computing professionals need ethics at all, what actually happens when someone acts unethically, why copyright legislation exists, how the different families of software licence differ (including the specific Free Software Foundation and Open Source Initiative philosophies named in the syllabus), and the social, economic and environmental impact of artificial intelligence, a topic the original notes on this page did not cover at all.
Why Computing Needs Ethics
A civil engineer's mistakes are visible in a building. A computing professional's mistakes are often invisible until they cause harm at enormous scale, a single line of code can affect millions of people simultaneously, and a system's failure mode is rarely obvious just from looking at it. That gap between "technically working" and "actually safe or fair" is exactly what ethics has to fill.
Of individuals whose data they handle, code that collects more data than it needs, or stores it insecurely, harms real people even if it never gets "hacked".
Of systems used in critical infrastructure, medical devices, vehicle control software and power grid systems fail in ways that directly threaten human life.
Of networks and personal information, a professional who cuts corners on security is choosing convenience over other people's safety.
Through automation, AI and misinformation, decisions made by a small team of engineers can reshape employment, public discourse and trust at a societal scale.
Professional Bodies: BCS and IEEE
The syllabus specifically expects you to understand why joining a professional ethical body matters. These bodies exist precisely because "act ethically" is too vague on its own, they turn it into an enforceable, specific code that members are held to.
- The UK's chartered professional body for computing
- Sets standards for computing professionals
- Members must follow a formal Code of Conduct
- Code covers: act in the public interest, maintain professional competence, respect confidentiality, work within the law
- A global professional body, US-based
- Develops technical standards used worldwide
- IEEE Code of Ethics covers: public safety, avoiding conflicts of interest, honesty, fairness to colleagues
Acting Ethically: Impact and Consequences
The syllabus explicitly requires you to reason about the impact of acting ethically or unethically for a given situation, not just recognise right from wrong in the abstract. Exam scenarios describe a specific dilemma and ask you to weigh what happens down each path.
Try it: work through real ethical dilemmas
Click a scenario to reveal what each path actually costs, and which professional code it touches.
Ethical path: disclose the flaw to the employer and affected users promptly, even though this is embarrassing and costly, because BCS's "act in the public interest" principle outweighs short-term reputational comfort. Users can then patch, change passwords or take precautions.
Unethical path: stay silent to protect the company's image or avoid the cost of a fix. This leaves real users exposed to a known, exploitable risk they have no way to protect themselves against, and if discovered later, causes far greater reputational and legal damage than early disclosure would have.
Ethical concern: a backdoor is a deliberate security weakness. Even if the stated intent is legitimate law enforcement, it can be discovered and exploited by anyone, undermining the security of every user of that software, not just the intended surveillance targets.
The professional's dilemma: this pits legal instruction against professional principle. Both BCS and IEEE codes prioritise public safety and the public interest, a professional weighing this has to consider whether they can refuse, escalate through professional or legal channels, or must resign rather than build it, exactly the kind of judgement call a code of conduct exists to guide.
In favour: employers have a legitimate interest in productivity and preventing data leaks or misuse of company systems, and can argue this is proportionate on company-owned equipment during work hours.
Against: constant, total surveillance is disproportionate to most legitimate goals, damages trust, and can capture far more personal information than necessary (personal messages, health-related searches, union organising). The ethical question is rarely "monitor or not" but "is the scope, transparency and proportionality of this specific monitoring justified", exactly the kind of nuanced judgement examiners reward over a flat yes or no.
Ethical path: suspend the tool until the bias is understood and corrected, even though this slows hiring and costs money, because deploying a system known to discriminate breaches the "fairness" and "public interest" principles in both codes, and may also break anti-discrimination law.
Unethical path: keep using it because it is fast and "not deliberately" biased. Not knowing why an AI system discriminates is not a defence, a professional's duty to understand the systems they deploy is itself part of acting ethically, this is a core reason AI's social impact is now an explicit syllabus topic.
Intellectual Property and Copyright
| Protection type | What it covers | Duration | Example |
|---|---|---|---|
| Copyright | Automatically protects creative works: software code, written text, music, images | 70 years after the author's death | The source code of an application |
| Patent | Protects inventions and innovations, requires formal registration | Up to 20 years | A novel piece of hardware |
| Trade secret | Confidential business information kept private | Indefinite, while it stays secret | Google's search ranking algorithm |
| Trademark | Protects brand names and logos | Renewable indefinitely | The Apple logo, the name "Windows" |
Why copyright legislation is needed
The syllabus specifically asks you to explain the need for copyright legislation, not just describe what it does. Without a legal right to control and profit from their own work, creators lose the financial incentive to invest time and money creating it in the first place.
- Anyone could copy and resell software the moment it is released, undercutting the original creator
- Developers and companies would have no financial incentive to invest in original work
- No legal mechanism would exist to stop large-scale commercial piracy
- Creators can license their work on their own terms, and be paid for it
- Legal recourse exists against unauthorised copying or distribution
- A functioning software industry becomes possible, because investment is protected
Try it: how long does each protection actually last?
Drag the slider to a year in the future and see which protections for a piece of work created today would still be active.
Copyright Infringement
Try it: step through what happens when infringement is discovered
Copyright holders generally have two separate routes available to them, and they are not mutually exclusive. Click through the steps below.
Software Licensing
The syllabus names four categories of licence you must be able to distinguish, and specifically expects you to know that "open source" itself splits into two different philosophies represented by two different real organisations: the Free Software Foundation (FSF) and the Open Source Initiative (OSI).
Pay to use, source code hidden, redistribution prohibited. Example: Microsoft Office, Adobe Photoshop.
Free to try for a limited time or with limited features, pay to unlock the full version. Example: a WinRAR trial.
The two faces of open source
The FSF's philosophy, expressed through licences like the GPL, insists that freedom must be preserved forward: if you distribute a modified version, you must also release your source code under the same licence. This is called copyleft, it uses copyright law to guarantee openness stays openness.
The OSI certifies licences like MIT and Apache that are far less restrictive: you can take the code, modify it, and release your version as closed-source commercial software with no obligation to share your changes back.
- Free to use and distribute
- A large community finds and fixes bugs fast
- Transparent, so security can be independently audited
- Customisable for specific needs
- No vendor lock-in
- No guaranteed dedicated support
- Anyone can create a malicious fork
- May lack a polished user experience
- Documentation is sometimes poor
- Commercial-grade support still costs money
Try it: which licence fits this situation?
Answer the three questions and the tool will recommend, and justify, a licence type, exactly the kind of "justify the use of a licence for a given situation" reasoning the syllabus asks for.
What Is Artificial Intelligence?
The syllabus requires you to understand both the applications of AI and its wider impact. Applications first, since the impact only makes sense once you know what the systems actually do.
Try it: explore where AI is actually used
Click a domain to see what AI does there, and the specific concern that comes with it.
Impact of AI
The syllabus is specific: you must understand the impact of AI across social, economic and environmental issues. Exam answers that only discuss "AI takes jobs" are answering a fraction of what is expected.
- Bias: AI trained on biased historical data can reproduce and scale up existing discrimination
- Deepfakes and misinformation, undermining trust in what people see and hear online
- Erosion of privacy through pervasive facial recognition and behavioural prediction
- Positive: accessibility tools, such as real-time captioning or image description for disabled users
- Job displacement in roles built on repetitive or predictable tasks
- New jobs and entire industries created around building, training and auditing AI systems
- Productivity gains for businesses that adopt AI effectively
- Market concentration, since training the largest models is only affordable for a small number of very large companies
- Training large AI models consumes enormous amounts of electricity and water for cooling data centres
- Growing e-waste from the specialised hardware AI training requires
- Positive: AI-optimised smart energy grids and precision agriculture can reduce waste elsewhere
Try it: classify the impact
Each statement below describes a real consequence of AI. Classify it as social, economic or environmental before revealing the answer.
Impact of Computing on Society
Zooming out from AI specifically, computing as a whole reshapes society across every one of these areas at once, usually with both a clear benefit and a genuine cost attached.
| Area | Positive impacts | Negative / ethical concerns |
|---|---|---|
| Employment | New tech jobs, remote working, productivity tools | Automation displaces manual workers, widening skills gap |
| Privacy | Better authentication, encrypted communications | Mass surveillance, data harvesting by corporations |
| Health | Medical diagnosis AI, wearable health monitors | Sedentary lifestyle from screen time, cyber-addiction |
| Environment | Smart energy grids, precision agriculture, remote work reduces travel | Data centre energy consumption, e-waste disposal |
| Education | Online learning, access to global knowledge | Digital divide, misinformation online |
| Artificial Intelligence | Medical imaging, fraud detection, accessibility tools | Bias in AI decisions, deepfakes, autonomous weapons |
Practice Questions
Professional bodies publish a formal code of conduct that turns vague expectations into specific, checkable standards, and members who breach it can be formally struck off, creating a real professional cost for unethical behaviour. This gives employers and clients a shared, trustworthy expectation of a member's conduct before working with them.
Without legal protection, anyone could copy and resell software immediately after release, undercutting the original creator and removing any financial incentive to invest time and money in developing original work. Copyright legislation gives creators a legal right to control and profit from their work and legal recourse against unauthorised copying, making continued investment in software viable.
The FSF promotes copyleft licences, such as the GPL, which require that any distributed modified version must also be released as open source under the same licence, preserving openness forward. The OSI certifies permissive licences, such as MIT and Apache, which allow modified versions to be released as closed-source commercial software with no obligation to share changes back.
An open source licence is appropriate, since it is free to use and its published source code can be studied and modified to fit the business's exact workflow. A permissive (OSI-style) licence specifically would also let the business keep any modifications private if it later wanted to, whereas a copyleft (FSF-style) licence would require sharing modifications back publicly.
AI's impact spans genuinely separate categories: social (bias in decisions, deepfakes, privacy erosion, but also new accessibility tools), economic (job displacement in some roles alongside entirely new industries and productivity gains), and environmental (the large energy and water cost of training big models, offset in part by AI-optimised systems like smart grids). A complete answer has to address more than one of these categories, since a system can be beneficial in one and harmful in another simultaneously.
This is copyright infringement. The developer could face a civil lawsuit from the original copyright holder seeking financial damages, and, for commercial-scale infringement, criminal prosecution. If done through an organisation, the business itself could also face fines, reputational damage, and in some jurisdictions the loss of its trading licence.
