Ethical Harms: Unauthorized Access and Manipulation of Societal Attitudes (CIPT: Privacy Technologist's Role in the Organization)
Introduction
In the IAPP CIPT (Certified Information Privacy Technologist) Body of Knowledge, privacy technologists must understand more than legal compliance. They must also recognize the ethical harms that can arise from how technology collects, processes and uses personal information. Two of the most heavily tested ethical harms are:
1. Unauthorized access to personal data and systems.
2. Manipulation of societal attitudes, meaning the use of data and technology to shape the beliefs, opinions and behaviors of individuals or whole populations.
These harms often occur even when an organization is technically "compliant." That is why the CIPT places them within the privacy technologist's role: technologists build the systems that either enable or prevent these harms.
Why It Is Important
- Harm goes beyond the law: Laws trail behind technology. An action can be legal but still unethical, such as micro-targeted political persuasion or dark patterns. Technologists are often the first to see these risks.
- Trust and reputation: Breaches and manipulation scandals cause lasting damage to brand trust and invite regulatory scrutiny. Cambridge Analytica and Facebook is the classic example of manipulation. Large breaches such as Equifax illustrate unauthorized access.
- Societal scale: Manipulation does not just harm one person. It can distort elections, deepen polarization, spread disinformation and undermine democratic discourse.
- Privacy by Design: The CIPT expects technologists to embed safeguards at the design stage rather than react after harm occurs.
- Frameworks: The topic connects to Solove's Taxonomy of Privacy, Calo's subjective and objective harms, NIST's Privacy Framework (problematic data actions) and the FIPPs. All of these appear on the exam.
What It Is
1. Unauthorized Access
Unauthorized access occurs when a person or system gains entry to personal data, devices or systems without permission or beyond the scope of their permission. It includes:
- External attacks: hacking, phishing, credential stuffing, malware and ransomware.
- Insider threats: employees browsing records out of curiosity (for example, hospital staff viewing a celebrity's medical file), misuse of privileged accounts, or data theft by departing staff.
- Scope creep: authorized users or third parties accessing data for purposes beyond what was agreed. This overlaps with secondary use.
- Physical access: stolen laptops, unsecured paper records and shoulder surfing.
Why it is an ethical harm and not just a security issue:
- It violates individuals' reasonable expectations and their autonomy over their information.
- It can lead to identity theft, discrimination, stalking, blackmail, embarrassment and emotional distress.
- In Solove's taxonomy, it maps to Information Processing harms (insecurity) and Invasions (intrusion).
- Calo would classify the fear of being watched as a subjective harm and actual misuse as an objective harm.
2. Manipulation of Societal Attitudes
This harm involves using personal data, algorithms and platform design to influence how people think, feel and act. It often happens without their awareness, and it operates at both individual and collective levels. Examples:
- Micro-targeting and psychographic profiling: building personality profiles from data such as social media likes to deliver tailored persuasive messages, as in political campaigns.
- Filter bubbles and echo chambers: recommendation algorithms optimized for engagement show users only content that reinforces existing views, increasing polarization.
- Disinformation and amplification: bots, fake accounts and algorithmic boosting that spread false narratives.
- Dark patterns: user interface designs that nudge people into choices they would not otherwise make. Examples include confirmshaming, hidden opt-outs, pre-ticked boxes and forced continuity.
- Behavioral nudging at scale: experiments such as Facebook's 2014 emotional contagion study, which altered users' news feeds to change their moods without informed consent.
- Discriminatory targeting: showing or withholding opportunities (housing, jobs, credit ads) based on inferred characteristics, which reinforces societal bias.
Why it is an ethical harm:
- It undermines autonomy: people's capacity to make free, informed decisions.
- It exploits information asymmetry and power imbalance between organizations and individuals.
- It threatens democratic processes and social cohesion.
- Solove's taxonomy classifies related behavior as Decisional Interference (an invasion) and also involves Aggregation and Secondary Use.
How It Works (Mechanisms and Lifecycle)
Unauthorized access pathway:
1. A weak control exists: excessive privileges, unpatched systems, poor authentication or no logging.
2. An actor exploits it: an external attacker, an insider or a third-party vendor.
3. Data is viewed, copied, altered or exfiltrated.
4. Harm materializes for individuals (fraud, exposure) and for the organization (fines, notification costs, loss of trust).
Manipulation pathway:
1. Collection: extensive behavioral data is gathered, often through tracking, cookies, SDKs and data brokers.
2. Aggregation and inference: data is combined to infer sensitive traits such as political leaning, mental state or sexual orientation.
3. Profiling and segmentation: people are grouped by susceptibility.
4. Targeted delivery: tailored content, ads or interface nudges are pushed, often opaquely.
5. Feedback loop: engagement data refines the model, which amplifies the effect over time.
Controls the Privacy Technologist Should Apply
Against unauthorized access:
- Least privilege and role-based access control (RBAC) or attribute-based access control (ABAC).
- Strong authentication: multi-factor authentication (MFA) and privileged access management.
- Encryption at rest and in transit, plus tokenization and pseudonymization.
- Logging, monitoring, audit trails, and data loss prevention (DLP).
- Data minimization and retention limits, because less data means less exposure.
- Vendor due diligence and contractual controls.
- Security awareness training and an incident response plan.
Against manipulation:
- Privacy by Design and ethics reviews, including privacy impact assessments (PIAs/DPIAs) that consider societal impact.
- Transparency: clear notices and explanations of why content or ads are shown.
- Meaningful consent and avoidance of dark patterns, with privacy-protective defaults.
- Purpose limitation, which prevents data collected for one reason from being repurposed for persuasion.
- Algorithmic fairness testing, bias audits and human oversight.
- Limits on sensitive inferences and on micro-targeting of vulnerable groups.
- User controls: opt-outs from personalization and access to profiling data.
- Value-sensitive design and ethics frameworks such as the IEEE guidelines, alongside regulatory principles like GDPR Article 22 on automated decision-making.
Key Frameworks to Link
- Solove's Taxonomy: four groups of harm: Information Collection, Information Processing, Information Dissemination and Invasions. Unauthorized access falls under insecurity and intrusion. Manipulation falls under decisional interference, aggregation and secondary use.
- Calo's Harm Dimensions: subjective harm (the perception of unwanted observation) versus objective harm (actual adverse use of information).
- NIST Privacy Framework: problematic data actions and their impact on individuals.
- FIPPs/OECD principles: security safeguards, purpose specification, use limitation, openness and individual participation.
- Contextual Integrity (Nissenbaum): information flows should match the norms of the context. Manipulation and unauthorized access both violate those norms.
Exam Tips: Answering Questions on Ethical Harms: Unauthorized Access and Manipulation of Societal Attitudes
1. Distinguish legal from ethical. If a scenario says something is legal but users would be surprised or upset, the exam is testing ethical harm. Do not choose "no issue because it is compliant."
2. Identify the harm type first. Ask whether someone is getting data they should not have (unauthorized access) or whether data is being used to influence beliefs or behavior (manipulation). The right control depends on this.
3. Map to Solove. Snooping employees and hacking point to insecurity or intrusion. Nudging or psychographic targeting points to decisional interference, aggregation or secondary use. Expect at least one taxonomy-mapping question.
4. Prefer preventive, design-level controls. When asked for the BEST action, favor Privacy by Design responses such as least privilege, data minimization, DPIAs and privacy-protective defaults over after-the-fact fixes.
5. Insider threats count. Unauthorized access is not only external hacking. An authorized employee viewing records without a business need is unauthorized access. Look for RBAC, need-to-know controls and audit logging as answers.
6. Recognize dark patterns. Pre-checked boxes, guilt-tripping language, buried opt-outs and asymmetric choice buttons are manipulation. The correct remedy is neutral, balanced choice architecture and clear consent.
7. Watch for societal scale. Answers that mention polarization, elections, disinformation or discrimination against groups signal "manipulation of societal attitudes," not just individual harm.
8. Remember the classic cases. Cambridge Analytica illustrates manipulation and secondary use. The Facebook emotional contagion study illustrates manipulation without consent. Healthcare snooping incidents illustrate insider unauthorized access.
9. Eliminate extreme answers. Options like "stop all data collection" or "ignore because users accepted the terms" are usually wrong. The CIPT favors balanced, risk-based approaches.
10. Consider the technologist's role. Questions may ask what the privacy technologist should do. Good answers include raising concerns, conducting a PIA, recommending technical controls, collaborating with legal and ethics teams, and documenting decisions.
11. Use key words as clues. "Without user knowledge," "inferred," "profiling," "nudge" and "engagement optimization" point to manipulation. "Credentials," "privileges," "breach," "accessed records" and "exfiltration" point to unauthorized access.
12. Link to autonomy and trust. When asked why something is harmful, strong answers reference loss of individual autonomy, violated expectations (contextual integrity) and erosion of trust.
Sample Question
A retail app redesigns its consent screen so that "Accept All" is a large, bright button while "Manage Preferences" is small grey text. Which ethical harm is MOST evident?
A. Unauthorized access
B. Manipulation through dark patterns
C. Data breach
D. Aggregation
Answer: B. The interface design steers users' decisions, which is decisional interference. The fix is balanced choice design.
Summary
Unauthorized access is about who can reach data. Manipulation of societal attitudes is about how data is used to shape minds and behavior. Both are ethical harms that a privacy technologist must anticipate and mitigate through design, technical controls, transparency and governance. On the exam, identify the harm, map it to a framework, and choose the most preventive, proportionate and user-respecting control.
