UPI Study › Courses › Principles of Management › Free lessons › Lesson 17.2 Free with a free account
Ethics in the Age of Tech: Navigating Digital Dilemmas
Principles of Management · Chapter 17: Ethics, Social Responsibility, and SustainabilitySection 17.2 · lesson 112 of 11713 min · free with a free account
Free preview
Create a free account to keep watching
It takes a minute and costs nothing. Your free account unlocks all of Chapter 1 and 3 more lessons of Principles of Management.
Lesson complete
Earn the 3 college credits for Principles of Management
Enrolling adds a quiz after each lesson, the graded work and the official transcript. Recommended for credit by NCCRS.
Next lesson: Building an Ethical Workplace: Culture, Codes, and Conduct →← Ethics vs. Social Responsibility: What Businesses Owe and to WhomNext lesson: Building an Ethical Workplace: Culture, Codes, and Conduct →
In this lesson
Navigating digital dilemmas.
- Identify the ethical questions raised by business technology
- Explain the governance needed around data and algorithms
- Recognize how bias enters an algorithmic system
- Assess a described technology proposal for its ethical dimension
Transcript
Everything said in the video, section by section (1,661 words). Read along, or search it.
Ethics, Social Responsibility, and Sustainability
Ethics in the Age of Tech
A new technology often arrives before anyone has agreed on the rules for using it. This section is about what a manager does in that gap. We'll start with data ethics, the difference between what an organization may do with data and what people expect. We'll look at primary and secondary use, the limits of consent, and why de-identification isn't absolute. Then algorithmic decisions. How bias gets in, why accuracy isn't fairness, and who is accountable when a model is wrong. We'll set out the contested question of automation and jobs, at full strength on both sides. We'll cover intellectual property, security, and the governance architecture these questions require. And we'll finish with a lucrative data partnership that a governance review stopped before it started.
New Situations, Old Principles
Lawful Use and Expected Use
Start with the general point that governs this whole section. Technology doesn't create new ethical principles. It creates new situations, in which existing principles must be applied. Those situations often arrive faster than law or social norms can adapt. So the frameworks from Subchapter seventeen point one still govern. What changes is the speed and novelty of the situations a manager must apply them to. Frequently, that happens well before any settled legal or social consensus has formed. Data ethics shows this clearly. Its practical question is a gap. On one side is what an organization may lawfully do with data. On the other is what the people it was collected from would actually expect. Lawful use and reasonably expected use are not the same question. An organization that operates entirely within the first, while ignoring the second, risks genuine ethical failure, even with no legal violation.
Data Use and Consent
Primary Use, Secondary Use, and Consent's Limits
Two terms organize the rest of data ethics. Primary use is data used for the purpose it was originally collected for. Secondary use is data used for a different purpose than the one originally stated. This is where consent runs into real limits. Consent given at collection doesn't stretch to a use nobody contemplated at the time. Agreeing to one stated purpose isn't agreeing to every use an organization might later find profitable. De-identification means removing directly identifying information from a dataset. It deserves a genuine caution. It isn't absolute. De-identified data can sometimes be re-identified, when it's combined with other available information. So de-identification reduces risk, without eliminating it entirely. The same question applies to data about employees, covered in Subchapter thirteen point seven, and to data about customers. Whichever population it concerns, does actual use match reasonable expectation?
Figure 17.2.1
From Collection to Third-Party Use
Figure seventeen point two point one follows data through Harborline Manufacturing, from collection to third-party use. At collection, the consent basis is the stated purpose. Primary use needs only one governance check. Stay within the organization's own stated policy. Secondary use moves further from that purpose, so it requires approval. Third-party use moves furthest of all, and it requires senior sign-off. Read left to right, and the governance requirement escalates at every transfer. That escalation isn't bureaucracy for its own sake. It reflects how far each transfer has moved from the purpose the original consent actually covered. The further data travels from its stated purpose, the more senior the person who should answer for it.
Algorithmic Decision Making
How Bias Gets In
Algorithmic decision making means using a model or automated system to make, or inform, a business decision. It's used across hiring screening, credit and pricing decisions, content recommendation, and operational scheduling, among other applications. Several ethical questions follow directly. The first is how bias gets in. Bias can enter through training data. The historical data used to build a model may reflect past patterns of unfair treatment. Bias can also enter through the definition of the outcome being predicted. Sometimes the target a model is trained to predict is an imperfect proxy, standing in for what the organization actually cares about. Then the model can predict the proxy well, and still miss the point. Notice that neither route requires anyone to intend unfairness.
Fairness and Accountability
Accurate Is Not the Same as Fair
The second question is subtler. A technically well-performing model can still allocate unfairly. A model can achieve strong accuracy, by the measure it was optimized for. And it can still produce systematically unequal outcomes across different groups. That's because accuracy and fairness are simply not the same measure. The third question is accountability. Suppose a recommendation is wrong, and causes harm. Who is actually responsible? The people who built the algorithm? The people who deployed it? Or the people who acted on its recommendation without question? Finally, there's a recurring tension between a model's performance and the ability to explain its output. Some of the most accurate modeling approaches are also among the hardest to explain. A person affected by the decision may not be able to understand the explanation at all. That's a real trade-off, between predictive power and the transparency an affected person might reasonably expect.
Table 17.2.1, Part 1
Governing Data Use and Algorithms
Table seventeen point two point one turns these questions into governance, for five common technology practices. For each, it names the ethical question, the control, and who should approve it. Using customer data for a new marketing purpose raises the expectation question. Does the new use match what customers reasonably expected at collection? The control is a secondary-use review against the original consent basis, approved by the data governance function. Deploying a hiring screening algorithm raises the bias question. Does the model's training data, or its target definition, embed unfair bias? The control is a bias audit before deployment, plus ongoing monitoring. Approval belongs to a cross-functional review that includes legal and HR. Sharing data with a third-party analytics partner asks whether the original consent ever contemplated this use. That needs senior sign-off with documented justification, from a senior executive accountable for the decision.
Table 17.2.1, Part 2
Governing Automation and Storage
Now the last two practices in the table. Automating a role previously performed by employees raises the question of obligation. What does the organization owe the workforce it displaces? The control is workforce transition planning, done before deployment. Senior leadership approves it, jointly with HR. Storing sensitive personal information raises a protection question. Is the information adequately protected against unauthorized access? The control is a security review and access controls, approved by the security or information governance function. Notice that every row ends with a named approver. That's what turns an ethical question into a decision someone is actually accountable for.
Automation and Employment
A Contested Question, Both Sides
Automation and employment raise a genuinely contested question. Both sides are presented at full strength, with no position taken. One position holds that displacement from automation is a normal, historically recurring feature of technological change. On this view, the gains from automation ultimately spread broadly. They arrive through lower costs, and through new roles the technology itself creates. So an organization's obligation extends to reasonable transition support. It doesn't extend to preserving roles the technology has made unnecessary. The opposing position holds that the pace and scale of contemporary automation differ meaningfully from past transitions. It argues that the gains disproportionately favor capital, over the displaced workforce. And it holds that the organization bears a genuine, more extensive obligation to the workers whose roles it automates. Why? Because the organization chose the automation, and captured most of its benefit. Notice that both positions accept some obligation. They disagree about how far it extends.
Intellectual Property and Security
2 More Governance Questions
Intellectual property raises questions of attribution, and of the use of generated content. These are described here at the level of the question, not the current legal position. First, what does an organization owe in attribution, when it uses content produced with substantial technological assistance? Second, what rights does it actually hold in the result? The legal position genuinely varies by jurisdiction, and it continues to develop. So this course doesn't state it. Any real situation requires current legal counsel. Security is the obligation to protect information held about others. It's best understood as a governance duty, rather than a purely technical concern. An organization that collects information about employees, customers, or others has taken on a genuine responsibility to protect it. And that responsibility belongs with organizational governance, not only with a technical security team working in isolation.
Governance Architecture
4 Elements, and the Pace Problem
What governance architecture do these questions require? Four elements. First, a clear decision on who approves a data use. Second, a review process for algorithmic systems, both before and after deployment. Third, documented accountability, naming who is actually responsible for a system's outcomes. And fourth, a route for someone affected to challenge an outcome they believe was wrong. One structural problem runs through this whole section. Call it the pace problem. Technology decisions are frequently made by people without the authority to consider their ethical dimension. Picture an engineer, or a product manager, implementing a technically sound solution. She may have neither the mandate nor the organizational support to pause, and examine the ethical questions. So the fix is structural. Build the governance architecture into the organization itself. Simply asking individuals to exercise authority the organization never gave them won't work.
Case Vignette
An Attractive Arrangement Built on Repurposed Data
Here's a composite case, constructed for instruction. Harborline Manufacturing was offered a lucrative analytics partnership by an outside firm. The partner proposed combining Harborline's customer service records with its own marketing datasets. The aim was to build more targeted advertising products. But those service records had been collected solely to resolve service issues. Before considering any commercial terms, Harborline's data governance function asked three questions. What consent basis had actually been obtained, when the service records were collected? Had customers contemplated this secondary, third-party use in any way at the time? And who in the organization held the authority to approve it, given how far it sat from the data's original purpose? The review found that the original consent language covered only service resolution. No reasonable customer expectation extended to third-party marketing use. Harborline declined the arrangement. Not because it was clearly unlawful, but because governance had surfaced a gap. It was the gap between lawful possibility and reasonable expectation, and it surfaced before any commercial decision was made. Next, Subchapter seventeen point three turns to the broader architecture that builds an ethical workplace generally.
About Principles of Management at UPI Study
Principles of Management (Business 101) is a self-paced online course of 117 video lessons in 17 chapters, from the classical schools and the four management functions to leadership, motivation, change, human resources, strategy and ethics. It carries a NCCRS credit recommendation for 3 semester credits.
Every lesson transcript is free on pages like this one, and the videos are free with a free UPI Study account. Enrolling adds a practice quiz after each lesson, the graded work and final exam that earn the credit, lifetime access, and an official transcript on completion. UPI Study has helped over 130,000 students with college credits. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
Enroll now → Course details Download the syllabus Every course with PRO →
Why take it here
Why students take Principles of Management with UPI Study
Where the credit goes
Universities that work with NCCRS
Principles of Management carries a NCCRS credit recommendation. These universities work with NCCRS; some are UPI Study's signed partners. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
Thomas Edison State UniversitySigned partner
University of the PeopleSigned partner
Charter Oak State CollegeSigned partner
Strayer UniversitySigned partner
Columbia Southern UniversitySigned partner
Indiana TechSigned partner
Penn StateWorks with NCCRS
Southern New Hampshire UniversityWorks with NCCRS
SUNYWorks with NCCRS
+many more cooperating universitiesFind my college →Find my college → Not listed? Ask Athena in the corner of this page, or book a 15-minute call.
FAQ
Questions students ask us
About Ethics in the Age of Tech: Navigating Digital Dilemmas
What is the difference between primary and secondary use of customer data?
How does bias get into algorithms used in business decisions?
Can an accurate algorithm still be unfair?
What governance does a company need for data and AI decisions?
About the Principles of Management course
How many college credits is Principles of Management?
How much does Principles of Management cost and what is included?
Which universities accept UPI Study credits?
How long does Principles of Management take?
How are the exams done?
How do my credits reach my college?
Ready for the full course? Earn 3 credits in Principles of Management.
117 video lessons, a quiz after each one, and the graded work that earns the credit. Recommended for credit by NCCRS.
3 credits · NCCRSEnroll →