2025 IB DP Digital society 模擬試題連答案詳解

Centralized digital identity systems without independent oversight create several human rights challenges. First, they enable mass surveillance by allowing the state to track all digital transactions and public services accessed by a citizen. Second, they can lead to discrimination and systemic exclusion, where marginalized populations without access to digital technology are locked out of critical services like healthcare or social welfare.

Award 1 mark for identifying the first threat (such as privacy loss or surveillance). Award 1 mark for identifying the second threat (such as exclusion, discrimination, or lack of recourse). Award 0.5 marks for linking the identification to the lack of independent oversight or state-control.

Autonomous delivery robots utilize sensors to safely navigate sidewalks. Common options include LiDAR, cameras, and ultrasonic sensors. LiDAR emits thousands of laser pulses per second, measuring the time they take to bounce back to construct a precise 3D representation of the environment. This helps the robot identify stationary obstacles like benches and moving obstacles like pedestrians to adjust its path dynamically.

Award 1 mark for identifying the first sensor (such as LiDAR, Ultrasonic, Radar, or Cameras). Award 1 mark for identifying the second sensor. Award 0.5 marks for outlining how the chosen sensor detects obstacles or measures distances.

When recruitment AI exhibits bias, it compromises the ethical values of fairness (ensuring objective evaluation) and equality (providing equal opportunities without systemic bias). To mitigate this, developers can ensure that training datasets are representative of all demographic groups and actively audit the data to remove historical biases.

Award 1 mark for identifying the first ethical value (such as fairness, equality, justice, or transparency). Award 1 mark for identifying the second ethical value. Award 0.5 marks for identifying a valid mitigation strategy used during the training phase.

Rural telemedicine is often hindered by the digital divide. Barriers include technical factors like the lack of reliable broadband infrastructure, and human factors like low digital literacy among elderly patients. To resolve this, governments or telecom companies can fund the expansion of high-speed network infrastructure, such as fiber-optic networks or satellite internet access points, to rural health clinics and community hubs.

Award 1 mark for identifying the first barrier (such as network access, cost, or infrastructure). Award 1 mark for identifying the second barrier (such as digital literacy, technical support, or device availability). Award 0.5 marks for suggesting a relevant infrastructure improvement.

First, the robot relies on sensors (such as LiDAR, depth cameras, or ultrasonic sensors) to gather real-time data about its surroundings, detecting obstacles like pedestrians or street furniture. Second, its onboard computer processes this input to identify potential collisions and plan an adjusted route. Third, the computer sends commands to its actuators (such as electric motors and braking systems) to adjust its speed and steering, physically navigating around the obstacles safely.

Award [1 mark] for explaining how sensors gather real-time environmental data. Award [1 mark] for explaining how this data is processed to make navigational decisions. Award [1 mark] for explaining how actuators execute the physical movements (speed/steering) to safely navigate around obstacles.

An exclusive transition to online portals assumes all patients have equal access to digital systems. However, patients affected by the digital divide may lack reliable broadband internet or the digital literacy required to use these platforms. Consequently, these individuals are unable to schedule essential appointments or access virtual consultations, which leads to delayed diagnoses, missed treatments, and ultimately worse physical health outcomes.

Award [1 mark] for identifying a relevant aspect of the digital divide (e.g., lack of internet access or digital literacy). Award [1 mark] for explaining how this barrier prevents the patient from accessing the exclusive online portal. Award [1 mark] for explaining how this lack of access leads to poorer health outcomes (e.g., delayed treatment, missed diagnosis).

Automated resume screening software is often trained on historical hiring data, which may contain past human biases and systemic inequalities. The algorithm identifies and replicates these patterns, systematically penalizing resumes containing keywords associated with marginalized groups (e.g., specific demographic indicators). This compromises fairness because qualified candidates are rejected based on systemic bias rather than their individual merit and qualifications.

Award [1 mark] for identifying that historical biased data is used to train the algorithm. Award [1 mark] for explaining how the algorithm replicates these biased patterns to disadvantage specific demographic groups. Award [1 mark] for explaining how this compromises fairness by evaluating candidates on biased criteria rather than their actual merit.

Arguments in favor of AMR deployment: 1. Operational Efficiency: AMRs can operate continuously without fatigue, optimizing the delivery of critical pharmaceuticals and lab samples. This reduces delivery times and minimizes human error in internal logistics. 2. Optimizing Human Resources: By automating routine transport tasks, clinical staff (nurses, assistants) are freed from logistics, allowing them to spend more high-quality, direct care time with patients. Arguments against AMR deployment and associated risks: 1. Privacy and Security: Transporting sensitive patient files via AMRs poses data protection risks. If a robot is intercepted, tampered with, or experiences a system malfunction, unauthorized individuals could access confidential health records. 2. Erosion of Human-Centered Care: Replacing human couriers with robots removes opportunities for informal social interactions between staff and patients, which are vital for patient well-being in clinical settings. 3. Safety and Reliability: Navigation errors in crowded hospital corridors could lead to collisions with fragile patients or medical equipment, creating safety hazards. Conclusion/Evaluation: While AMRs offer significant logistical advantages, their implementation is only ethical and effective if accompanied by strict security safeguards, such as biometrically locked cargo compartments and encrypted communication. Ultimately, the decision is justified only if the hospital actively redirects the saved staff hours into direct, empathetic patient care, ensuring that technology enhances rather than diminishes the human element of healthcare.

Markband Descriptors: [1-2 marks]: Minimal or superficial understanding of the scenario. Identifies only basic advantages or disadvantages of robots without clear connection to healthcare or ethical concepts. [3-4 marks]: Explains some benefits (e.g., speed, efficiency) and challenges (e.g., cost, technical glitches) of AMRs. The analysis is mostly descriptive and lacks balance or depth in discussing privacy and human care. [5-6 marks]: Provides a balanced and detailed analysis of both the operational efficiencies of AMRs and the ethical/social implications (privacy, human-centered care). Relevant digital society concepts and terminology are used appropriately. [7-8 marks]: Offers a well-structured, balanced, and critical evaluation. Successfully synthesizes the tension between operational efficiency and patient well-being. Concludes with a clear, reasoned judgment or recommendation supported by the preceding analysis.

### Detailed Solution Outline

**1. Introduction**
- **Contextualization:** State-sponsored surveillance technologies (such as NSO Group's Pegasus spyware, facial recognition systems, and mass internet traffic monitoring) have given governments unprecedented power to monitor citizens.
- **Definitions:** Define 'international human rights frameworks' (e.g., Universal Declaration of Human Rights [UDHR] Article 12 on privacy and Article 19 on freedom of expression; International Covenant on Civil and Political Rights [ICCPR]).
- **Thesis Statement:** To a moderate extent. Human rights frameworks are vital for establishing ethical and legal benchmarks and empowering advocacy, but their real-world impact is heavily constrained by state sovereignty, weak enforcement mechanisms, and technological lag.

**2. Arguments Supporting the Efficacy of Human Rights Frameworks (To some extent)**
- **Setting Global Standards:** Treaties like the ICCPR establish clear legal standards of 'proportionality' and 'necessity' that state surveillance must meet. This limits the legal justification for arbitrary surveillance.
- **Legal Redress and Advocacy:** Human rights bodies (e.g., the European Court of Human Rights or UN Human Rights Committee) provide platforms for citizens and NGOs (like Citizen Lab and Amnesty International) to challenge state behavior, leading to landmark rulings that force policy changes.
- **Influence on National Legislation:** International frameworks guide the creation of national laws (e.g., GDPR in Europe, or national privacy acts) which translate abstract international rights into enforceable domestic regulations.
- **Diplomatic and Economic Pressure:** High-profile UN reports and investigations can name and shame non-compliant states and restrict the export of dual-use surveillance technologies to human rights violators.

**3. Arguments Highlighting Limitations of Human Rights Frameworks (To a limited extent)**
- **Sovereignty and Enforcement Deficit:** International law fundamentally relies on the consent of sovereign states. Powerful nations can ignore UN recommendations or international court rulings without facing direct physical or economic consequences.
- **National Security Exemptions:** Most international frameworks permit the suspension or limitation of certain rights during times of national emergency or for national security purposes, a loophole frequently exploited by authoritarian and democratic states alike to justify mass surveillance.
- **Jurisdictional Arbitrage:** The commercial surveillance industry operates globally. Companies can be headquartered in one country, develop code in another, and sell to a third, making it difficult for any single international framework to regulate their activities.
- **Technological Lag:** International treaty-making and consensus-building processes are slow, whereas surveillance technologies (e.g., generative AI for mass monitoring, zero-click exploits) evolve at an exponential rate, leaving existing regulatory frameworks outdated.

**4. Synthesis & Conclusion**
- Synthesize the tension between global human rights ideals and the realities of sovereign political power.
- Conclude that while international frameworks are indispensable as a moral compass and a tool for civil society, they must be paired with technical countermeasures (like end-to-end encryption), domestic judicial independence, and direct sanctions on the private surveillance market to effectively mitigate surveillance threats.

### Marking Rubric (12 Marks)

**Criterion A: Focus and structure (3 marks)**
- **3 marks:** The response is consistently focused on the prompt. There is a clear, coherent, and well-structured argument with an explicit thesis statement that directly addresses the 'to what extent' command term.
- **2 marks:** The response is mostly focused on the prompt and has a logical structure, though some arguments may lack development or drift into descriptive narrative.
- **1 mark:** The response lacks focus, structure, or a clear thesis.

**Criterion B: Knowledge and understanding of digital society (3 marks)**
- **3 marks:** Demonstrates excellent knowledge and understanding of relevant digital society concepts, such as state surveillance, privacy, freedom of expression, sovereignty, and international law, using precise terminology.
- **2 marks:** Demonstrates good knowledge of digital society concepts, though there may be minor inaccuracies or a lack of depth in explaining the connection to human rights frameworks.
- **1 mark:** Knowledge is superficial, descriptive, or contains significant errors.

**Criterion C: Analysis and evaluation (4 marks)**
- **4 marks:** Offers a balanced and critical analysis of both the strengths and limitations of international frameworks. Evaluates the tension between state power, technology, and international law with sophisticated synthesis and a well-reasoned conclusion.
- **3 marks:** Provides a balanced discussion of both sides, but the evaluation/synthesis is less developed or relies on generalized assertions.
- **2 marks:** The analysis is one-sided or mostly descriptive, focusing only on the limitations or only on the strengths of frameworks without critical evaluation.
- **1 mark:** Minimal or superficial analysis.

**Criterion D: Use of examples (2 marks)**
- **2 marks:** Employs relevant, specific, and contemporary real-world examples (e.g., Pegasus spyware, GDPR, UN declarations, specific court rulings) that directly support the argument.
- **1 mark:** Examples are used but are generic, outdated, or not fully integrated into the argument.
- **0 marks:** No relevant examples are provided.

### Detailed Solution Outline

**1. Introduction**
- **Contextualization:** State-sponsored surveillance technologies (such as NSO Group's Pegasus spyware, facial recognition systems, and mass internet traffic monitoring) have given governments unprecedented power to monitor citizens.
- **Definitions:** Define 'international human rights frameworks' (e.g., Universal Declaration of Human Rights [UDHR] Article 12 on privacy and Article 19 on freedom of expression; International Covenant on Civil and Political Rights [ICCPR]).
- **Thesis Statement:** To a moderate extent. Human rights frameworks are vital for establishing ethical and legal benchmarks and empowering advocacy, but their real-world impact is heavily constrained by state sovereignty, weak enforcement mechanisms, and technological lag.

**2. Arguments Supporting the Efficacy of Human Rights Frameworks (To some extent)**
- **Setting Global Standards:** Treaties like the ICCPR establish clear legal standards of 'proportionality' and 'necessity' that state surveillance must meet. This limits the legal justification for arbitrary surveillance.
- **Legal Redress and Advocacy:** Human rights bodies (e.g., the European Court of Human Rights or UN Human Rights Committee) provide platforms for citizens and NGOs (like Citizen Lab and Amnesty International) to challenge state behavior, leading to landmark rulings that force policy changes.
- **Influence on National Legislation:** International frameworks guide the creation of national laws (e.g., GDPR in Europe, or national privacy acts) which translate abstract international rights into enforceable domestic regulations.
- **Diplomatic and Economic Pressure:** High-profile UN reports and investigations can name and shame non-compliant states and restrict the export of dual-use surveillance technologies to human rights violators.

**3. Arguments Highlighting Limitations of Human Rights Frameworks (To a limited extent)**
- **Sovereignty and Enforcement Deficit:** International law fundamentally relies on the consent of sovereign states. Powerful nations can ignore UN recommendations or international court rulings without facing direct physical or economic consequences.
- **National Security Exemptions:** Most international frameworks permit the suspension or limitation of certain rights during times of national emergency or for national security purposes, a loophole frequently exploited by authoritarian and democratic states alike to justify mass surveillance.
- **Jurisdictional Arbitrage:** The commercial surveillance industry operates globally. Companies can be headquartered in one country, develop code in another, and sell to a third, making it difficult for any single international framework to regulate their activities.
- **Technological Lag:** International treaty-making and consensus-building processes are slow, whereas surveillance technologies (e.g., generative AI for mass monitoring, zero-click exploits) evolve at an exponential rate, leaving existing regulatory frameworks outdated.

**4. Synthesis & Conclusion**
- Synthesize the tension between global human rights ideals and the realities of sovereign political power.
- Conclude that while international frameworks are indispensable as a moral compass and a tool for civil society, they must be paired with technical countermeasures (like end-to-end encryption), domestic judicial independence, and direct sanctions on the private surveillance market to effectively mitigate surveillance threats.

### Marking Rubric (12 Marks)

**Criterion A: Focus and structure (3 marks)**
- **3 marks:** The response is consistently focused on the prompt. There is a clear, coherent, and well-structured argument with an explicit thesis statement that directly addresses the 'to what extent' command term.
- **2 marks:** The response is mostly focused on the prompt and has a logical structure, though some arguments may lack development or drift into descriptive narrative.
- **1 mark:** The response lacks focus, structure, or a clear thesis.

**Criterion B: Knowledge and understanding of digital society (3 marks)**
- **3 marks:** Demonstrates excellent knowledge and understanding of relevant digital society concepts, such as state surveillance, privacy, freedom of expression, sovereignty, and international law, using precise terminology.
- **2 marks:** Demonstrates good knowledge of digital society concepts, though there may be minor inaccuracies or a lack of depth in explaining the connection to human rights frameworks.
- **1 mark:** Knowledge is superficial, descriptive, or contains significant errors.

**Criterion C: Analysis and evaluation (4 marks)**
- **4 marks:** Offers a balanced and critical analysis of both the strengths and limitations of international frameworks. Evaluates the tension between state power, technology, and international law with sophisticated synthesis and a well-reasoned conclusion.
- **3 marks:** Provides a balanced discussion of both sides, but the evaluation/synthesis is less developed or relies on generalized assertions.
- **2 marks:** The analysis is one-sided or mostly descriptive, focusing only on the limitations or only on the strengths of frameworks without critical evaluation.
- **1 mark:** Minimal or superficial analysis.

**Criterion D: Use of examples (2 marks)**
- **2 marks:** Employs relevant, specific, and contemporary real-world examples (e.g., Pegasus spyware, GDPR, UN declarations, specific court rulings) that directly support the argument.
- **1 mark:** Examples are used but are generic, outdated, or not fully integrated into the argument.
- **0 marks:** No relevant examples are provided.

The deployment of automated facial recognition systems (FRT) in public places raises significant human rights concerns. First, the right to privacy is challenged because FRT continuously scans and processes the biometric data of individuals without their active consent. Second, the right to freedom of assembly and association is threatened, as the awareness of mass surveillance can have a chilling effect, discouraging individuals from participating in peaceful protests or public gatherings.

Award 1 mark for each valid human right identified, up to a maximum of 2 marks. Acceptable answers include: Right to privacy, Right to freedom of assembly/association, Right to non-discrimination (due to algorithmic bias against specific demographics), Right to freedom of movement. Do not award marks for vague answers like 'general fairness' without linking to a specific human right concept.

Integrating autonomous delivery drones into crowded cities presents clear ethical and social challenges. First, planners must ensure physical safety and accessibility, as drones operating on sidewalks or low heights can pose hazards to vulnerable pedestrians (like visually impaired individuals). Second, there is a risk of job displacement, where low-barrier delivery jobs are automated, potentially exacerbating local economic inequalities.

Award 1 mark for each valid ethical or social challenge identified, up to a maximum of 2 marks. Acceptable answers include: Physical safety of pedestrians, Accessibility/obstruction issues for disabled persons, Economic displacement/unemployment of manual delivery workers, Noise pollution impacting mental well-being, Unequal access to the delivery service across different socioeconomic neighborhoods.

The student explains a digital governance challenge related to algorithmic transparency in public systems. 1 mark: Identifies a relevant governance challenge (e.g., proprietary intellectual property, 'black box' machine learning models). 1 mark: Explains how this challenge manifests in the given context (e.g., hiding evaluation weights or criteria from the public). 1 mark: Connects the explanation to the wider governance impact (e.g., inability to appeal decisions, erosion of democratic accountability, or loss of public trust).

Award [1] for identifying a clear governance challenge related to algorithmic transparency (e.g., intellectual property restrictions, algorithmic complexity). Award [1] for explaining how this challenge functions within the decision-making system. Award [1] for linking this to a broader governance implication such as the erosion of public trust or the limitation of administrative appeal rights.

The student explains how the ethical concept of accountability is impacted by an algorithmic error. 1 mark: Identifies the degradation or diffusion of responsibility (e.g., the 'problem of many hands' or automation bias). 1 mark: Explains how this diffusion occurs between different stakeholders (e.g., government workers vs. software developers). 1 mark: Connects this to the ethical impact on the vulnerable families (e.g., lack of clear liability, inability to seek recourse, or systemic administrative injustice).

Award [1] for identifying how accountability is diffused or obscured (e.g., blaming the machine, automation bias, or distributed responsibility). Award [1] for explaining this dynamic within the context of the CivicPath system (e.g., bureaucrats deferring to automated outputs). Award [1] for explaining the consequences for affected stakeholders, such as the loss of human oversight or the prevention of meaningful administrative redress.

### Arguments for reliance on ADM systems (Benefits):
- **Administrative Efficiency and Scalability:** Automated systems can process thousands of applications in seconds, significantly reducing backlogs and freeing up human social workers to focus on complex cases.
- **Consistency and Objectivity:** Unlike human evaluators, who may suffer from fatigue or unconscious bias, an algorithm applies the exact same set of rules to every applicant, potentially reducing subjective discrimination.
- **Data-Driven Resource Optimization:** These systems can analyze vast data sets to identify trends, forecast demands, and ensure welfare resources are directed where they are statistically needed most.

### Arguments against reliance / Risks (Challenges to Governance and Human Rights):
- **Algorithmic Bias and Discrimination:** AI systems trained on historical data often inherit and perpetuate existing societal biases, disproportionately denying benefits to minority or marginalized groups.
- **Lack of Transparency ('Black Box' Problem):** Many advanced neural networks or proprietary algorithms do not offer explainable decisions. If an applicant is rejected, they have a right to know *why*, which is a fundamental tenet of administrative justice and human rights.
- **Erosion of Due Process and Accountability:** When a system makes a mistake, determining accountability (whether it lies with the software developer, the municipal authority, or the caseworker) is highly complex. Additionally, automated rejections may lack an easy, human-led appeals process.
- **Dehumanization of Social Support:** Welfare distribution requires empathy and context-specific understanding (e.g., sudden family emergencies) that static data fields and algorithms cannot comprehend.

### Conclusion / Evaluation ('To what extent'):
To protect human rights, governments must not rely fully on automated systems. Instead, they should adopt a hybrid model where systems like SmartAssist function exclusively as triage or advisory tools. The final decision to deny or reduce benefits must always be made by a human officer (human-in-the-loop). Furthermore, implementation must be contingent on strict governance frameworks: open-source transparency of the algorithm, mandatory human rights impact assessments (HRIAs), and regular independent audits to detect and mitigate bias.

### Markband Rubric (8 Marks)

**Level 1 (1–2 marks):**
- The response is largely descriptive and lists some basic points about computers or automated systems in government.
- Connection to human rights or administrative efficiency is superficial or absent.
- Minimal or no evaluation of the 'to what extent' aspect.

**Level 2 (3–5 marks):**
- Explains both benefits (e.g., speed, consistency) and drawbacks (e.g., bias, lack of empathy) of using ADMs like SmartAssist.
- Some connection is made to Digital Society concepts, specifically governance, transparency, or human rights.
- The analysis may be unbalanced, heavily favoring one side, or lacks a well-justified conclusion.

**Level 3 (6–8 marks):**
- Provides a balanced, sophisticated evaluation of the tension between administrative efficiency and the protection of human rights.
- Effectively uses relevant domain-specific terminology (e.g., algorithmic bias, black-box problem, human-in-the-loop, due process, accountability).
- Offers a clear, nuanced conclusion addressing 'to what extent' (e.g., proposing hybrid systems or regulatory safeguards).
- Directly addresses the context of governance and public administration challenges.

**Introduction**
Neo-Veridia's metropolitan authority is at a crossroad where it must balance collective security with individual human rights. This evaluation weighs Option A (centralized AI predictive policing) against Option B (decentralized, privacy-preserving community alerts) to formulate a sustainable policy recommendation.

**Evaluation of Option A (Centralized AI & Predictive Policing)**
* **Benefits**: Proactive security deployment, high efficiency in identifying wanted suspects in real-time, and deterrence of public crime due to constant surveillance. It leverages advanced data analytics to allocate scarce police resources efficiently.
* **Drawbacks**: Significant human rights issues regarding the right to privacy and anonymity in public spaces. Predictive algorithms are often trained on historical arrest data, which can perpetuate systemic biases against marginalized groups (algorithmic bias). It creates an asymmetrical dynamic of **power** where the state holds absolute surveillance capabilities over citizens, potentially chilling free speech and assembly.

**Evaluation of Option B (Decentralized, Privacy-Preserving Community Network)**
* **Benefits**: Minimizes the collection of personally identifiable information (PII), preserving the **identity** and privacy of citizens. It fosters community agency and democratic **governance** by giving residents control over local security. It avoids the algorithmic biases and false positives associated with facial recognition.
* **Drawbacks**: Highly dependent on active citizen participation and digital literacy, which could lead to uneven coverage or "vigilante" reporting. It is primarily reactive rather than predictive, and physical infrastructure upgrades require significant upfront capital expenditure without immediate analytical outputs.

**Recommendation and Justification**
It is recommended that Neo-Veridia implements **Option B**.

While Option A offers powerful analytical capabilities, its socio-technical risks—particularly systemic bias, erosion of democratic trust, and violations of article-level human rights (privacy and non-discrimination)—present too high of a cost to Neo-Veridia's democratic values. Option B aligns more closely with democratic **governance** frameworks that respect citizens' rights while addressing security concerns locally.

To address the limitations of Option B (such as latency or lack of participation), the metropolitan authority should:
1. Establish clear data governance protocols to prevent community alert platforms from becoming hubs for profiling or misinformation.
2. Ensure the physical security upgrades (e.g., better street lighting, emergency contact booths) are equitably distributed to prevent digital divide issues where only affluent areas benefit.

**Markband Descriptors (12 Marks Total)**

* **Level 4 (10–12 marks)**:
* The response provides a highly balanced, detailed, and critical evaluation of both Option A and Option B.
* The recommendation is persuasive, logical, and fully justified based on the preceding analysis.
* Digital society concepts (such as power, governance, and identity/human rights) are integrated naturally and accurately throughout the response.
* The candidate suggests robust, realistic mitigation strategies for the limitations of their chosen option.

* **Level 3 (7–9 marks)**:
* The response evaluates both options, but one may be treated with slightly more depth than the other.
* A clear recommendation is made and supported by arguments, though some links to digital society concepts could be more explicit.
* Relevant concepts are used correctly to support the evaluation.
* Some awareness of the limitations of the chosen option is demonstrated.

* **Level 2 (4–6 marks)**:
* The response describes both options but lacks deep, balanced critical evaluation (focuses mostly on benefits or drawbacks).
* The recommendation is present but may feel abrupt, simplistic, or weakly justified.
* Digital society terms are mentioned but not deeply integrated or explained.

* **Level 1 (1–3 marks)**:
* The response is highly descriptive, simply repeating information about the options.
* The recommendation is missing or merely stated without supporting arguments.
* No meaningful integration of digital society concepts or frameworks.

* **0 marks**: No rewardable content.