Applied ethics case studies: worked answers for technology and environment dilemmas

An applied-ethics case study is a short, realistic dilemma in which you, playing an officer, scientist or administrator, must choose a course of action where values collide, in technology, environment, medicine or public life. UPSC tests this format because the GS-4 paper reserves its heaviest weight, usually 20 marks and sometimes 25, for case studies: they reveal whether you can reason with values under pressure, not merely recite definitions. Applied-ethics cases are harder than administrative ones because the right answer is not written in any rulebook; you must weigh rights against consequences, certainty against precaution, and present a decision you can defend. This article works through six such dilemmas, each built from the scenario types the examiner returns to again and again: artificial intelligence and data privacy, genetically modified crops, misinformation on social media, clinical-trial ethics, industrial pollution, and environmental clearance for infrastructure.

How to work a case study

The framework article for this series gives a seven-step method for cracking any case study. First, open with a hook and a two-line summary that names the real conflict, not just the story. Second, set the compass: the two or three values that will govern every later judgment. Third, map the stakeholders and what each stands to win or lose. Fourth, list the ethical issues as issues, never as narration. Fifth, name the dilemma precisely and resolve it with a method, such as Kidder's four paradigms: truth versus loyalty, individual versus community, short term versus long term, and justice versus mercy. Sixth, evaluate each option with its merit, its demerit and its ethical reading. Seventh, commit to one course of action, add safeguards, and close with the values it serves.

The six cases below demonstrate every one of these moves in action.

Answer-writing pointers for applied-ethics answers

Applied-ethics cases are graded on distinctions, not declarations. Five habits from the detailed playbook's technology and current-affairs material:

• Define the technology in one line before judging it. The examiner cannot grade an ethical analysis of a tool you have not described accurately; precision first, principles second.
• Invoke the precautionary principle by name. Where harm is serious, irreversible or borne by the unconsenting, say so explicitly and place the burden of proof on those who would act.
• Split the fused categories. Therapy from enhancement, somatic from germline, reversible from irreversible, national from global: the marks live in the distinctions, and each half gets its own ethical reading.
• Think in three time horizons. The patient today, the institution in a decade, the species in a century; intergenerational equity is the signature value of this section, so let the long horizon appear in every answer.
• End with governance, not just a verdict. A registry, an oversight body, a published allocation formula, a review clause: new technologies need institutions, not only judgments.

Worked case studies

Read each case as an examiner would write it: a messy situation, a protagonist under pressure, and no clean answer. Work the moves on paper before reading the recommended course of action, because the habit of structured thinking is what the marks reward.

The facial recognition tender

The situation: You are the Police Commissioner of a large Indian metropolis. After a spate of street crimes and a stampede at a religious gathering, the state government sanctions a city-wide smart surveillance project built on AI-driven facial recognition to identify suspects in real time. The lowest bidder, a foreign firm, offers the system at a price no domestic vendor can match. But its model was trained almost entirely on non-Indian faces, and independent audits abroad found it misidentifies women and darker-skinned persons at far higher rates. The vendor refuses to disclose the training data, calling it a trade secret, and requires that footage of Indian citizens be uploaded to its overseas servers for "model improvement". The Home Minister wants the system inaugurated before a mega sporting event six weeks away. Civil society groups warn of mass surveillance and a chilling effect on protest. Your own officers are split: the crime branch wants the tool now, the legal cell fears wrongful arrests will destroy public trust for a decade.

• You, the Police Commissioner: the decision-maker, answerable for both crime control and citizens' liberties.
• City residents, especially women and minorities: bearers of the rights to privacy and equality, and the first victims of misidentification.
• Police personnel: want faster investigations but will carry the blame for wrongful arrests.
• The foreign vendor: seeks market entry and a stream of Indian training data.
• The Home Minister and state government: want visible, quick results before the mega-event.
• The courts: will ultimately judge the system's legality and proportionality.
• Civil society and the press: watchdogs against mission creep and data misuse.

Ethical issues and dilemmas: The central clash is public safety versus privacy and equality. A biased model converts a policing tool into a discrimination engine, offending equality before the law and the fundamental right to privacy recognised in the Puttaswamy judgment. Opacity compounds the harm: a black-box system cannot be questioned in court, violating due process. Data sovereignty is at stake when citizens' biometric data crosses borders for "model improvement". There is the dilemma of proportionality: is blanket surveillance of millions justified by crimes committed by a few? And the dilemma of mission creep: a system bought for suspect identification can quietly become a tool for tracking protesters and journalists.

Options before you: Four paths lie open, each with a price.

• Option 1: accept the tender as it stands. Merit: the system goes live before the mega-event, deterrence is immediate, and the price is the lowest. Demerit: biased misidentification, unconstitutional data transfer abroad, and the first wrongful arrest will ignite a crisis of legitimacy.
• Option 2: reject the tender outright. Merit: protects privacy and equality, and signals that rights are not for sale. Demerit: the city loses a genuine crime-fighting aid, and the project may simply be awarded by your successor with fewer safeguards.
• Option 3: conditional acceptance with safeguards. Merit: balances safety and rights through audits, local data storage and human oversight. Demerit: slower and costlier, and the vendor may withdraw rather than open its model.
• Option 4: a limited pilot of a domestic, audited system. Merit: builds evidence before scaling and supports indigenous capability. Demerit: will not be ready for the mega-event, and coverage stays limited.

Recommended course of action: Choose a strengthened version of Option 3, and if the vendor refuses the conditions, fall back to Option 4. Sign only after an independent bias audit on Indian faces, with results published; require all processing and storage on servers within India; mandate a human-in-the-loop rule so that no arrest rests on an algorithmic match alone; and write in a sunset clause and a public annual impact assessment so the system cannot quietly expand. This course honours proportionality: the state may restrict liberty for security, but only to the least extent necessary and with safeguards. It also protects your fiduciary duty to citizens, who trusted the police with their faces only insofar as the law permits. Record the Home Minister's pressure in the file noting, so the decision trail shows the choice was yours and the reasons were recorded.

Values invoked: Proportionality, privacy as a fundamental right, equality and non-discrimination, accountability, transparency, the precautionary principle, data sovereignty, and fidelity, the officer's loyalty to citizens over political convenience.

Safeguards and systemic measures: Beyond the immediate decision, build the institution. Create a standing technology-assessment cell in the police headquarters that must clear every algorithmic system before procurement, with a published checklist: bias audit, data localisation, human-in-the-loop, and sunset review. Route all future surveillance proposals through a civilian oversight committee so the decision is never one officer's lonely call. Publish the audit results and the annual impact assessment on the department's website; transparency is what converts a policing tool into a publicly legitimate one. And train investigating officers that an algorithmic lead is the beginning of an investigation, never its conclusion.

What the examiner rewards: The examiner rewards the refusal to be stampeded by the event deadline, the Puttaswamy and proportionality reasoning used as tools rather than ornaments, and a decision that protects both safety and liberty instead of sacrificing one. Bonus marks for the sunset clause: it shows you understand that power granted in an emergency rarely returns on its own.

The GM mustard file

The situation: You are the Chairperson of the country's apex regulatory committee for genetically modified crops. A dossier seeking commercial release of a high-yield GM mustard hybrid has been pending for years. The developer, a public university spin-off with private investors, claims the hybrid yields 25 to 30 per cent more than conventional varieties, could cut India's massive edible-oil import bill, and has passed every biosafety test the rules require. But the file is uneasy reading. Two field trials were conducted in seasons with unusually favourable weather; the raw data behind the "no gene flow" claim has not been shared, citing commercial confidentiality; and an internal note from your predecessor flags a whistleblower's allegation that adverse observations in one trial were quietly reclassified. Farmer unions are divided: one bloc demands the seeds before the next sowing season, another fears contamination of the organic and export market. Environmental groups warn that the herbicide-tolerance trait will lock farmers into heavier chemical use. The Agriculture Minister's office has "enquired" twice this month about the timeline.

• You and the committee: regulators on whose credibility the whole biosafety regime rests.
• Farmers: split between those who want higher yields and those who fear market and ecological costs.
• The developer company and its investors: want approval and first-mover returns.
• Consumers: eat the oil; their food safety is at stake.
• Environmental groups and organic exporters: guardians of biodiversity and premium markets.
• Future generations: inherit whatever genes escape into the wild.
• The whistleblower: risks a career to flag gaps in the data.

Ethical issues and dilemmas: The core dilemma is food security versus food safety under uncertainty, the classic home of the precautionary principle. Approving on incomplete data gambles with an irreversible release: unlike a drug that can be recalled, genes in the wild cannot be withdrawn. There is a conflict of interest when the developer funds the very science the regulator relies on, and a transparency deficit when raw data hides behind "commercial confidentiality". Distributive justice matters too: the yield gains may accrue to seed companies and large farmers while the ecological and market risks fall on smallholders and organic growers. Finally, there is the dilemma of regulatory capture: clearing the file under ministerial "enquiries" would convert a scientific decision into a political one.

Options before you: Four paths lie open, each with a price.

• Option 1: approve commercial release now. Merit: higher yields, lower imports, a pro-farmer headline. Demerit: irreversible biosafety risk on contested data; regulatory credibility collapses if harm surfaces.
• Option 2: reject the application permanently. Merit: zero biosafety risk. Demerit: forfeits genuine yield gains; history shows bans breed illegal, unmonitored seed spread, which is worse than regulated release.
• Option 3: conditional, staged approval. Merit: an evidence-based middle path with labelling, buffer zones and surveillance. Demerit: expensive, slow, and politically unpopular with all sides.
• Option 4: defer for independent re-testing with open data. Merit: restores trust through independent public science. Demerit: delays benefits by seasons; the developer may litigate.

Recommended course of action: Choose Option 4 first, with Option 3 as the pathway after it. Order independent re-testing of the contested trials by public agricultural universities, require the developer to place the complete raw dataset in the public domain as a condition of any approval, and commission a full-season, multi-location assessment including gene-flow studies. Only then consider staged release: limited agro-climatic zones, mandatory labelling, buffer and refuge requirements, and a post-release surveillance programme funded by the developer through an escrow. If independent testing confirms safety, approve; if not, reject with published reasons. This treats uncertainty as a reason to investigate, not a licence to gamble, and it answers the ministerial "enquiries" with a file note: the committee's duty is to evidence, not timelines.

Values invoked: Precautionary principle, scientific integrity, transparency, intergenerational equity, distributive justice, public trust in regulation, and the courage to resist political pressure.

Safeguards and systemic measures: Fix the process, not just the file. Make open data the default: every biosafety dossier must carry its raw trial data in the public domain, with commercial-confidentiality claims adjudicated by the committee, not asserted by the applicant. Institutionalise independent re-testing by public laboratories for every contested application, funded by an applicant fee held in escrow. Create a formal whistleblower-protection channel inside the regulatory system so the next flagged note reaches the chairperson directly. And publish reasoned orders for every approval and rejection; a regulator that explains itself earns the trust that a silent one cannot command.

What the examiner rewards: Marks flow to the candidate who treats uncertainty as a reason to investigate rather than to gamble, who spots the conflict of interest in industry-funded science, and who answers political pressure with a file note rather than a surrender. The escrow-funded surveillance shows you know that regulation is a continuing relationship, not a one-time signature.

The viral rumour

The situation: You are the District Magistrate of a communally sensitive district. At dawn, a video showing a mob attack on a place of worship begins circulating on messaging apps, captioned as happening in your district's old town. Your fact-check unit confirms within two hours that the clip is actually from another country, three years old, a textbook deepfake-style manipulation of context. But truth travels slower than outrage: crowds are gathering at two chowks, stone-pelting has begun near the bus stand, and rumour-mongers are forwarding the clip with calls for "retaliation". The Superintendent of Police urges a 48-hour blanket internet shutdown to break the viral chain. Traders, doctors and journalists oppose it: the district runs on digital payments, the civil hospital's telemedicine link and ambulance coordination depend on data networks, and board examinations are underway. Your political superiors hint that a shutdown would look "decisive". Meanwhile the debunk video your team posted has a tenth of the rumour's reach.

• You, the DM: responsible for both public order and citizens' rights.
• Residents of the tense localities: potential victims of violence, and of the shutdown's collateral damage.
• The police: need tools to prevent escalation.
• The fact-check unit and local journalists: truth-tellers with limited reach.
• Traders, hospitals, students: bear the shutdown's economic and human cost.
• The rumour's originators: seek to engineer a riot.
• Political superiors: want visible decisiveness.

Ethical issues and dilemmas: The dilemma is public order versus free expression and livelihood, to be resolved by proportionality, not panic. A blanket shutdown is a blunt instrument: it punishes the entire district for the crimes of a few forwarders, violating the least-restrictive-means test the Supreme Court has laid down for internet restrictions. There is also the dilemma of speed: the rumour's virality outpaces official rebuttals, and silence from the administration reads as confirmation. A deeper issue is trust: citizens who distrust official channels will believe the rumour over the debunk. Finally, there is the temptation of political signalling: choosing the shutdown to appear decisive serves the superior's image, not the district's safety.

Options before you: Four paths lie open, each with a price.

• Option 1: a 48-hour blanket internet shutdown. Merit: instantly breaks the forwarding chain and signals strong action. Demerit: cripples payments, hospitals and examinations; disproportionate; rumours migrate offline and return angrier.
• Option 2: no restrictions, only counter-messaging. Merit: preserves rights and normal life. Demerit: may be too slow against a viral rumour; risks violence the administration could have prevented.
• Option 3: a targeted, surgical response. Merit: proportionate and rights-respecting, combining limited forwarding curbs with amplified rebuttals. Demerit: demands tight coordination and may still leak at the edges.
• Option 4: shutdown plus mass arrests of forwarders. Merit: maximum deterrence. Demerit: punishes panic-sharing alongside malice; arrests without due process inflame the situation.

Recommended course of action: Choose Option 3, executed with visible authority. Impose a temporary, geographically limited restriction on bulk forwarding and new-group creation rather than a district-wide blackout; flood every channel with the debunk, voiced not by the administration alone but by trusted local religious leaders, mohalla committees and influencers whose word carries further than a press note; deploy visible, restrained police presence at the chowks; open a 24-hour rumour helpline and announce it by loudspeaker; and register cases against the originators under the IT Act while distinguishing malicious creators from frightened forwarders. Place the proportionality reasoning on record, citing the Court's least-restrictive-means standard, so the file shows why a blanket shutdown was considered and rejected. Decisiveness is measured in outcomes, not in how many services you switch off.

Values invoked: Proportionality, least-restrictive means, the duty of truth-telling, public order, trust-building, accountability of malicious actors, and impartiality toward all communities.

Safeguards and systemic measures: Turn this crisis into capacity. Institutionalise a district rapid-response protocol: a verified fact-check unit with pre-authorised access to community radio, loudspeaker networks and influencer contacts, so the debunk launches within the hour, not the day. Run regular digital-literacy drills with mohalla committees, training citizens to check before forwarding. Keep a written, court-tested proportionality template for any future restriction proposal, so the least-restrictive-means analysis is a habit, not an improvisation. And after every incident, publish a short after-action note: what spread, what worked, what failed.

What the examiner rewards: The examiner rewards proportionality argued with the Court's own standard, the distinction between malicious creators and frightened forwarders, and the insight that trusted local voices carry the debunk further than any press note. Choosing visible authority over a blackout shows judgment, not weakness.

The trial that cut corners

The situation: You are Dr. Rao, a senior scientist leading a 40-member team at a reputed Indian pharmaceutical company. You are months from completing trials of a drug for a fast-spreading viral fever, and the pressure is crushing: investors are restless, the share price slides with every delay, and the board has already announced a launch date to the media. In a closed-door meeting, two senior colleagues propose "pragmatic" shortcuts: drop from the final dataset the trial participants who suffered adverse reactions, publish only the positive outcomes, skip the lengthy informed-consent process for illiterate volunteers by letting field workers sign on their behalf, and quietly use a molecule patented by a rival, betting the rival will settle rather than sue. They argue that every week's delay costs lives and that perfect trials are a luxury India cannot afford. The drug regulator, they remind you, has always trusted your company's dossiers. Your junior researchers are watching you for a signal.

• You, Dr. Rao: the scientist whose signature certifies the data.
• Trial volunteers, many poor and illiterate: trust the team with their bodies.
• Future patients: will take the drug on the strength of your data.
• The company, its board and investors: want speed and returns.
• The regulator and the public: rely on honest dossiers.
• The rival company: owner of the patented molecule.
• Junior researchers: are learning what science permits.

Ethical issues and dilemmas: The clash is beneficence against non-maleficence: the urge to rush a cure versus the duty not to harm. Dropping adverse reactions is data fraud, and a drug approved on laundered data can injure millions, a harm no launch-date arithmetic can justify. Informed consent is the moral core of human research: having field workers sign for illiterate volunteers treats the poor as raw material, violating autonomy and dignity. Using the rival's patented molecule adds theft to the fraud. There is also the dilemma of role responsibility: as team leader, your silence would teach forty young scientists that integrity is negotiable under pressure.

Options before you: Four paths lie open, each with a price.

• Option 1: accept the shortcuts. Merit: fastest launch, commercial survival, possibly early relief for patients. Demerit: unsafe-drug risk, fraud, betrayal of volunteers, and if exposed, the reputation of the company and of Indian pharma collapses.
• Option 2: quietly resign and walk away. Merit: keeps your hands clean. Demerit: abandons the volunteers and juniors to the fraud; the drug still launches.
• Option 3: refuse, document, and escalate through institutions. Merit: protects patients and integrity while using proper channels. Demerit: delays the project, invites retaliation, and the board may sideline you.
• Option 4: blow the whistle publicly at once. Merit: maximum sunlight and the fastest external intervention. Demerit: bypasses internal remedies and may kill a genuinely promising drug along with the fraud.

Recommended course of action: Choose Option 3, with Option 4 held in reserve. State your refusal in the meeting and record it in writing; submit the complete, unedited dataset to the company's ethics committee and the regulator; insist on a genuine informed-consent process with independent witnesses for illiterate volunteers; and require the rival's molecule to be licensed properly or replaced. If the company buries your report or proceeds with the fraud, escalate to public whistleblowing, because your fiduciary duty to the volunteers and future patients outweighs loyalty to the employer. Tell the juniors plainly why you refused: a scientist's signature is a promise to strangers who will swallow the pill. Speed that is bought with fraud is not beneficence; it is a slower, larger harm wearing the mask of urgency.

Values invoked: Non-maleficence, autonomy through informed consent, veracity and scientific integrity, fiduciary duty to trial participants, courage, and the Helsinki principle that the volunteer's welfare comes before the research.

Safeguards and systemic measures: Protect the next trial before it begins. Strengthen the institutional ethics committee with genuinely independent members and the power to halt a trial, not merely advise. Require that informed consent for illiterate volunteers be taken before an independent witness with audio-visual recording, as the rules already intend. Build a protected internal-reporting channel so a junior researcher can flag data manipulation without going through the very seniors proposing it. And rotate the data-audit function to an external statistician; when the people who generate the data also grade it, fraud becomes a temptation.

What the examiner rewards: This answer earns marks for naming data fraud plainly, for placing the volunteer's autonomy above the company's timeline, and for sequencing the response: refuse, document, escalate, and only then go public. The line to the juniors shows the examiner you understand that a leader's real product is the culture left behind.

The river of solvents

The situation: You are the Regional Officer of the State Pollution Control Board in a fast-growing industrial belt. A cluster of e-waste recycling units employs nearly four thousand migrant workers, but for years it has discharged untreated effluent, rich in lead, cadmium and solvents, into the river that irrigates downstream farms and feeds the town's drinking-water intake. An NGO-commissioned health survey found elevated heavy-metal levels in children of the riverside village; the industry association calls the study methodologically flawed and warns that closures will push the workers into starvation and simply shift recycling to unregulated backyard furnaces. The cluster's largest owner belongs to a politically influential family. Your junior officer's inspection report, which documented midnight dumping, has gone "missing" from the file twice. Farmers downstream are threatening a highway blockade, and a local journalist has asked for your on-record comment.

• You, the Regional Officer: the regulator, answerable to law and conscience.
• Riverside villagers and downstream farmers: bear the health and livelihood costs.
• Migrant workers: depend on the units for survival.
• Unit owners, including the politically connected family: profit from weak enforcement.
• The NGO and the journalist: demand accountability.
• The Board's chairman: must back your action or bury it.
• Future residents: inherit the poisoned aquifer.

Ethical issues and dilemmas: The central dilemma is the right to health versus the right to livelihood, sharpened by environmental injustice: the poorest drink the poison while the profits leave the district. The precautionary principle applies even with contested evidence, because heavy metals in children's blood is not a hypothesis to debate endlessly but a harm to stop. There is the dilemma of regulatory capture: missing reports and political ownership test whether the Board regulates industry or serves it. Sealing the units overnight would be a different injustice, destroying four thousand livelihoods, while doing nothing makes you complicit in slow poisoning.

Options before you: Four paths lie open, each with a price.

• Option 1: seal all units immediately. Merit: stops the discharge at once and sends a strong deterrent signal. Demerit: livelihood shock for thousands; recycling migrates underground where it is worse.
• Option 2: do nothing, citing inconclusive evidence. Merit: no disruption, no political friction. Demerit: continued poisoning and complicity; the missing reports implicate you.
• Option 3: time-bound compliance with enforcement teeth. Merit: addresses health and livelihood together through treatment plants, health screening and published reports. Demerit: needs monitoring capacity and political backing; visible results are slower.
• Option 4: relocate the cluster to a compliant industrial park. Merit: the cleanest long-term solution. Demerit: takes years; workers lose income in the transition.

Recommended course of action: Choose Option 3, and treat the missing reports as a separate integrity battle. Issue closure directions that take effect on a fixed, short deadline unless the units connect to a common effluent treatment plant funded through an escrow they cannot touch; order immediate health screening and treatment for the village children at the cluster's cost under the polluter-pays principle; publish every inspection report online so the next "missing" file is impossible; and escalate the tampering to the Board chairman in writing, seeking protection for your junior officer. Pair enforcement with a transition plan: skilling support and phased compliance so workers are not sacrificed for the owners' violations. Regulation that destroys livelihoods loses legitimacy; regulation that spares polluters loses its soul. The middle path is not weakness here, it is the only course that serves both justice and prudence.

Values invoked: Environmental justice, the precautionary principle, the polluter-pays principle, intergenerational equity, integrity and courage against regulatory capture, procedural fairness, and compassion for workers.

Safeguards and systemic measures: Make compliance the default, not the exception. Move the Board to continuous online effluent monitoring with public dashboards, so midnight dumping shows up on a screen instead of in a missing file. Digitise every inspection report with tamper-evident logs and automatic escalation when a report is altered or deleted. Create a formal protection protocol for field officers who report interference, because regulatory capture begins by intimidating the junior. And fund a worker-transition corpus from penalties collected, so the next enforcement drive does not force a choice between health and hunger.

What the examiner rewards: Marks reward the refusal of both lazy extremes, seal-everything and do-nothing, the polluter-pays reasoning, and the separate integrity battle over the missing reports. Pairing enforcement with worker transition shows you see the people behind the pollution, on both sides.

The highway through the tiger corridor

The situation: You are the Director (Projects) in the Ministry of Environment, Forest and Climate Change, handling environmental clearances. A flagship expressway that would cut travel time between two state capitals needs a 22-kilometre stretch through a notified tiger corridor, the last genetic link between two breeding populations. The environmental impact assessment was conducted only in the dry season, when wildlife movement is lowest; the Wildlife Institute's dissent note was kept off the appraisal committee's agenda. The state government, with elections approaching, presses for clearance "with standard conditions" such as elevated sections and underpasses. Conservationists warn that such conditions are routinely diluted after clearance and that this corridor, once severed, cannot be restored. Tribal families living in the corridor fear displacement without proper settlement of their rights under the Forest Rights Act. Your Minister wants the file "by Friday".

• You, the Director: custodian of a decision that outlives governments.
• The tiger populations and the corridor ecology: cannot speak, but have moral standing.
• Tribal families: face displacement and rights violations.
• The state government and the Minister: want speed and visible development.
• Commuters and businesses: genuine beneficiaries of connectivity.
• Conservationists and the Wildlife Institute: the scientific conscience of the process.
• Future generations: inherit either a connected forest or a severed one.

Ethical issues and dilemmas: The dilemma is development versus conservation, but the sharper issue is procedural integrity: a dry-season EIA and a suppressed dissent note mean the decision would rest on manufactured certainty, violating the precautionary principle. Intergenerational equity forbids trading an irreplaceable corridor for a travel-time saving, since extinction and severance are irreversible. The tribal question adds distributive justice: the Forest Rights Act requires their claims to be settled before any diversion, and "Friday" cannot override statute. Finally, there is the dilemma of obedience versus conscience: the Minister's timeline is a political preference, while your duty is to the law and the evidence.

Options before you: Four paths lie open, each with a price.

• Option 1: clear the project with standard conditions. Merit: connectivity, jobs, political goodwill. Demerit: likely irreversible ecological damage; conditions are historically diluted; rewards a flawed EIA.
• Option 2: reject the clearance outright. Merit: protects the corridor decisively. Demerit: real economic and political costs; may be reversed on appeal without better evidence.
• Option 3: return the proposal for a supplementary assessment. Merit: evidence-based, legally robust, and protective of both interests. Demerit: delays the project and invites political displeasure.
• Option 4: approve a realigned route avoiding the corridor. Merit: both the road and the forest survive. Demerit: longer route, higher cost, land-acquisition delays.

Recommended course of action: Choose Option 3, with Option 4 as the preferred outcome of the reassessment. Return the file with written reasons: a full-season, multi-year EIA, the Wildlife Institute's dissent placed on the committee's record, verified settlement of Forest Rights Act claims, and a funded, monitored mitigation design, with elevated corridor sections paid for by the project, not promised as an afterthought. Recommend the standing committee consider realignment as the first alternative. Record the Minister's "by Friday" direction in the file noting alongside your reasons for needing more time: obedience to a superior does not extend to signing an inadequately assessed clearance. A highway can be built next year; a severed gene pool cannot be rebuilt in any year.

Values invoked: Intergenerational equity, the precautionary principle, integrity and professional courage, procedural justice, trusteeship of nature, respect for tribal rights, and fidelity to statute over political timelines.

Safeguards and systemic measures: Reform the clearance pipeline itself. Require full-season, multi-year EIAs as a non-negotiable minimum, with the terms of reference published before the study begins. Give dissenting scientific institutions a guaranteed place on the appraisal agenda, with their notes annexed to the final order. Make Forest Rights Act compliance a certified precondition, verified by the gram sabha, not a post-clearance promise. And publish every clearance with its conditions, timelines and compliance reports in a searchable public database, so diluted conditions become visible to courts and citizens alike.

What the examiner rewards: The examiner rewards spotting the procedural fraud before the ecological one, the insistence that Friday cannot override the Forest Rights Act, and the realignment recommendation that refuses the false binary of road versus forest. Recording the Minister's direction in the file shows institutional courage, not defiance.

Just War Theory: the ethics of organised violence

Applied ethics is not only about technology and the environment; it also asks whether organised violence can ever be morally bounded. Just War Theory is the tradition, originating with St. Augustine and St. Thomas Aquinas, that assesses both the resort to war and conduct within it. It divides into two halves, and a mains answer should test any conflict against both, never just one.

Jus ad bellum is the justice of going to war: the set of conditions that must hold before war itself is justified, including legitimate authority, just cause, right intention, a reasonable probability of success, last resort, and proportionality of ends. Jus in bello is justice in the conduct of war: the rules that bind even a belligerent whose cause is just, because a just cause does not license unjust means.

Three principles govern conduct within war. Discrimination is the principle that combatants must be distinguished from non-combatants, and that non-combatants must never be deliberately targeted. Proportionality is the principle that the harm inflicted must not be excessive in relation to the military advantage gained. Military necessity is the principle that force may be used only so far as it is necessary to achieve legitimate military objectives, and no further.

Modern conflict strains the theory. Hybrid warfare blends cyberattacks, sanctions, propaganda and proxy forces, blurring the line between war and diplomacy; nuclear deterrence pushes full wars into limited and proxy forms; and non-state actors sit outside the state-centric frame entirely. Powerful states also face standing ethical charges: selective humanitarianism, intervening where interests lie while ignoring comparable suffering elsewhere; arms transfers that deepen civilian harm; sanctions that punish populations more than regimes; and delayed diplomacy held back for strategic leverage. Against these, the theory's answer is multilateralism, impartial aid, and accountability through international law.

The welfare algorithm

The situation: You are the district collector. The state has piloted an artificial intelligence system that ranks households for a welfare housing scheme, promising speed and zero favouritism. Field workers report that the system is rejecting eligible tribal families at three times the rate of others; the vendor calls the model 'proprietary' and will not explain individual decisions. The state government wants a statewide rollout before the next review meeting.

• The rejected families, denied a roof without being told why.
• You, answerable for decisions you cannot explain.
• The vendor, guarding a trade secret built on public data.
• The state government, wanting scale and speed.
• The public exchequer, funding a scheme that must be fair to be legitimate.

Ethical issues and dilemmas: The clash is between administrative efficiency and fairness, and four concerns from AI ethics bite at once. AI ethics is the set of moral principles guiding the responsible design and deployment of artificial intelligence. Discrimination: the model has learned the biases in its training data, so it punishes the already marginalised, which is algorithmic discrimination, unfair treatment produced or amplified by an automated system's biased data or design. Explainability: a decision that cannot be traced cannot be justified or appealed. Responsibility: when the algorithm errs, the vendor blames the data and the state blames the vendor, leaving the citizen with nobody to hold accountable. Data privacy: the system ingests household data whose consent and safeguards are unclear.

Options before you: Speed is tempting; scale makes the bias permanent.

• Option A: Roll out statewide as directed. Merit: speed, and political credit for modernisation. Demerit: you scale an unexamined bias to lakhs of families.
• Option B: Pause the rollout, publish the rejection patterns, require the vendor to open the model to independent audit, and build a time-bound human appeal for every rejection before any expansion. Merit: fairness is restored and accountability fixed. Demerit: delay, and friction with the state and the vendor.
• Option C: Roll out with manual review of rejections layered on top. Merit: a human safety net. Demerit: reviewers rubber-stamp at scale, and the bias persists underneath.

Recommended course of action: Choose B. No welfare decision may rest on an unexplainable black box: commission an independent audit, require the vendor to disclose the decision logic (trade secrecy cannot outweigh citizens' rights), publish aggregate rejection data by social group, and guarantee every rejected family a human appeal within a fixed time. Deploy the system as a recommender under human sign-off, never as an unreviewable decider.

Values invoked: Justice, because a housing scheme must not digitise discrimination. Accountability, because every rejection must have an author who can be questioned. Transparency, because public money and public data demand public reasons.

Two more worked cases from the detailed playbook

Two further applied-ethics cases, adapted from the detailed ethics playbook, apply the same seven moves to the global and the technological: a pandemic queue and an edited embryo.

The vaccine queue

The situation: You are a senior official in the health ministry of a country that manufactures vaccines. Production is limited and domestic demand is not yet met, yet several poorer nations that cannot manufacture doses plead for supply, and a global facility requests a share at cost. The domestic media demands our people first; your epidemiologists warn that uncontrolled spread abroad will return as new variants that defeat today's vaccines.

• Your citizens: waiting for doses; their government owes them protection first.
• The requesting nations: without manufacturing capacity, their people have no other source.
• The manufacturer: caught between domestic orders and global appeals.
• Global health bodies: allocating by need across borders.
• Future populations everywhere: who will face the variants bred in unvaccinated millions.

Ethical issues and dilemmas: The core is individual versus community at the level of nations: national partiality against global impartiality. There is also short term versus long term: domestic politics rewards our people first today, while epidemiology warns that no one is safe until everyone is safer. And there is equity versus ability to pay: whether life-saving doses flow to need or to purchasing power.

Options before you: Three allocations, and the extremes both fail a different duty.

• Option 1: export nothing until domestic demand is fully met. Merit: politically safe; every dose stays home. Demerit: variants bred abroad return to defeat the domestic programme; a moral failure with epidemiological interest charged on it.
• Option 2: export freely while domestic demand waits. Merit: maximal global equity. Demerit: domestic deaths that the government could have prevented; the political collapse of the programme helps no one.
• Option 3: tiered, rules-based allocation. Merit: defensible both at home and abroad; need, not noise, decides. Demerit: pleases no extreme; both the nationalists and the idealists will call it betrayal.

Recommended course of action: Choose Option 3. Secure the domestic vulnerable first: the elderly, health workers and the high-risk, by transparent priority. Then honour a fixed, published export quota to the global facility and to bilateral requests assessed on need, with the allocation formula open to scrutiny and no backdoor deals. A pandemic is the textbook case where the national and the global interest converge over time: unvaccinated populations anywhere breed the variants that defeat vaccination everywhere, so enlightened self-interest and solidarity prescribe the same policy.

Values invoked: distributive justice, enlightened self-interest, transparency, and solidarity: no one is safe until everyone is safer.

Safeguards and systemic measures: Publish the allocation formula and a real-time dashboard of doses produced, delivered and exported. Commission an independent audit of the queue. Invest, starting now, in distributed manufacturing partnerships so that the next pandemic does not begin with the same bottleneck.

What the examiner rewards: The examiner rewards the candidate who sees that our people first and the world first are the same policy on a longer timeline, and who refuses to let doses flow to purchasing power rather than need.

The edited embryo

The situation: You chair the national bioethics committee. A research hospital seeks permission for a first-in-country trial editing the genes of human embryos to eliminate a devastating inherited disease. This is germline editing: the changes would pass to every future generation of that line. The science is promising, but off-target effects are poorly understood; meanwhile, desperate families are petitioning for approval, and researchers warn that a refusal will push the work to jurisdictions with no oversight at all.

• The affected families: living with a devastating disease; hope is not an abstraction to them.
• Future children and generations: the most affected party, and the only one that cannot consent.
• The researchers: genuinely seeking cures, genuinely seeking prestige.
• You, the regulator: writing a rule that will outlive the headlines.
• Society: whose shared genome is the thing being edited.

Ethical issues and dilemmas: The first issue is beneficence versus non-maleficence: the hope of curing against the risk of harming. The second is consent: the generations who will inherit the edit cannot be asked. The third is irreversibility: an off-target effect, once inherited, cannot be recalled. The fourth is the slide from therapy to enhancement, and the fifth is justice: if the wealthy edit first, the human genome becomes another inequality.

Options before you: Three regulatory postures, from rush to freeze.

• Option 1: approve the clinical trial. Merit: hope for suffering families; scientific leadership. Demerit: irreversible, unconsented changes to the human germline on the basis of poorly understood risks.
• Option 2: ban all germline research permanently. Merit: maximum safety; no inherited mistakes. Demerit: freezes a technology that may one day cure; drives the research underground, beyond any oversight.
• Option 3: moratorium on clinical use with regulated basic research. Merit: honours both caution and hope; keeps the science visible and governed. Demerit: the waiting families pay the price of patience.

Recommended course of action: Choose Option 3. The precautionary principle governs exactly this terrain: where the harm is serious, irreversible and borne by the unconsenting, the burden of proof lies on those who would act, not on those who would wait. Permit laboratory research under strict oversight so understanding grows; require long-term safety data and genuine public deliberation before any clinical step; keep somatic therapies, which affect only the treated individual, on a separate and faster track. The families' suffering is real, and the answer to it is faster safe science, not unsafe haste.

Values invoked: the precautionary principle, intergenerational equity, autonomy (including the autonomy of the unborn), and scientific integrity: the courage to say not yet.

Safeguards and systemic measures: Create a national registry of all gene-editing research with mandatory reporting. Require independent ethics review with lay members for every germline protocol. Draw the legal line, in statute, between somatic therapy and germline modification. Coordinate internationally so that a refusal at home does not become an experiment abroad.

What the examiner rewards: Marks flow to the candidate who distinguishes somatic from germline editing without being asked, who names the unconsenting future generations as stakeholders, and who refuses the false choice between banning science and rushing it.

Key Terms

• Applied ethics: the application of moral principles to concrete, real-world dilemmas in fields such as technology, environment and medicine.
• Precautionary principle: when an action risks serious or irreversible harm, uncertainty is a reason to pause and investigate, not a licence to proceed.
• Proportionality: a restriction on rights is justified only if it is suitable, necessary, and the least intrusive means of achieving a legitimate aim.
• Informed consent: a person's voluntary agreement to an intervention, given with full understanding of its risks and alternatives; the moral core of research and medical ethics.
• Algorithmic bias: systematic, unfair error in an automated system that disadvantages particular groups, often inherited from unrepresentative training data.
• Data sovereignty: the principle that a nation's citizens' data should be stored and governed under its own laws rather than surrendered to foreign servers.
• Environmental justice: the fair distribution of environmental benefits and burdens, so that the poor do not disproportionately bear pollution's costs.
• Polluter-pays principle: the party responsible for pollution bears the cost of managing it and remedying the damage.
• Intergenerational equity: the obligation to preserve natural and institutional resources so future generations inherit undiminished opportunities.
• Kidder's paradigms: four classic dilemma frames, truth versus loyalty, individual versus community, short term versus long term, and justice versus mercy, used to name the real conflict in a case.
• Fiduciary duty: the obligation of one entrusted with another's interests, such as an officer to citizens or a scientist to trial volunteers, to act solely in their best interests.
• Distributive justice: fairness in how the benefits and burdens of a decision are shared across groups, so that the vulnerable do not carry costs created for others' gains.
• Germline editing: genetic modification of embryos or reproductive cells whose effects are inherited by future generations; distinguished from somatic editing, which affects only the treated individual.

Practice questions

You are the Food Safety Commissioner of a state. A laboratory report finds that a bestselling children's snack, made by the state's largest food employer, contains lead marginally above the permissible limit. The company calls it a testing anomaly and offers its own laboratory's clean report; it warns that a recall during the festival season will idle ten thousand workers. Parents' groups demand an immediate ban, while the industry minister asks you to "handle it quietly". The company's public-relations team has already planted stories questioning the public laboratory's competence.

• Question 1: Identify the ethical issues, including the conflict between the company's private test and the public laboratory's finding.
• Question 2: Lay out your course of action, from verification to public communication.

Framing hint: a precautionary-principle archetype; children's health is a non-negotiable, so verify independently, act in proportion to the risk, and communicate before rumour does.

You are the District Education Officer. An audio clip allegedly capturing a senior teacher "confessing" to leaking board-examination papers is spreading among parents. A forensic expert tells you privately that the voice shows signs of AI manipulation, but a full analysis needs a week. The teacher denies the charge and is distraught; parents demand suspension; a local leader wants the teacher made an example of. Meanwhile, a parents' messaging group has begun circulating the teacher's home address.

• Question 1: What are the ethical issues when the evidence itself may be manufactured?
• Question 2: What do you do in the next 72 hours to protect both the examination's integrity and the teacher's dignity?

Framing hint: tests due process in the deepfake age: verification before punishment, interim measures that do not prejudge, and the duty to correct the record publicly once the facts are in.

You are the Municipal Commissioner of a city choking on a mountain of legacy waste. A private firm offers a waste-to-energy plant that will clear the landfill in three years, but its emissions data comes from a differently designed plant abroad, and residents of the adjoining low-income colony fear dioxin exposure. The firm promises jobs and "European standards"; the residents' association threatens to block the project. The state pollution board's own monitoring station near the colony has been non-functional for eight months.

• Question 1: Map the stakeholders and name the environmental-justice dimension.
• Question 2: Decide, with conditions or refusal, and design the monitoring regime you would impose.

Framing hint: development promises versus lived exposure; independent verification of the claims, continuous public emission data, and binding penalties turn a promise into accountability.

You are the Collector of a drought-hit district. The upstream reservoir holds barely enough water for either the city's drinking supply through summer or one protective irrigation release that would save the standing crop of forty thousand small farmers. The city's legislators demand the water be reserved for taps; farmer unions warn of suicides if the crop fails; the reservoir engineer says splitting it risks losing both. The meteorological department forecasts a late monsoon, with no significant inflow expected for ten weeks.

• Question 1: Name the dilemma and the competing claims of justice at stake.
• Question 2: Choose and defend an allocation, including the relief measures for the side that loses.

Framing hint: a distributive-justice triage: separate the non-negotiable minimum (drinking water for survival) from the compensable loss (the crop), and pair the decision with transparent criteria and relief.

You are the Director of the state forensic science laboratory. In a high-profile custodial-death case, senior police officers press you to "confirm" a DNA match before the validation runs are complete, hinting that your laboratory's pending grant depends on cooperation. The victim's family and the press are watching, and your junior analysts look to you. A television channel has announced a prime-time segment tonight on the laboratory's "delay tactics".

• Question 1: Identify the ethical issues, including the threat to scientific integrity and the conflict of interest in the grant.
• Question 2: State your response to the officers and the safeguards you would institutionalise.

Framing hint: scientific integrity is not negotiable under pressure: no result before validation, written requisitions only, and an audit trail that protects both the innocent and the accused.

You are the district education officer. Parents complain that schoolchildren are spending hours on immersive metaverse platforms: cases of online bullying, contact with strangers, and falling grades. The platform's operator, a large technology company, points to parental controls and refuses to share safety data, calling it proprietary.

• Question 1: Identify the ethical issues, including the vulnerability of children, data privacy, and the platform's responsibility toward young users.
• Question 2: What regulatory and educational measures would you recommend, balancing innovation with child safety?

Framing hint: a playground that cannot show its safety record is not a playground; children's attention is not inventory.