Ethics for Computer Science & Engineering

Spring 2025
Vincenzo D'Andrea | James Brusseau
vincenzo.dandrea@unitn.it | jbrusseau@pace.edu
http://cse.ethicsworkshop.org/
Mondays, A109 17:30 - 19:30 p.m. 
Tuesdays, A107 15:30 - 17:30 p.m. 
UniTrento website

 

 

AI Ethics
Principles

Values

Autonomy/Freedom

  • Giving rules to oneself. (Between slavery of living by others' rules, and chaos of life without rules.) Kant.
  • Freedom maximization / Freedom limited only by freedom. (Do what you want up to the point where it interferes with others doing what they want.) Locke.
  • Practical independence: Physically free (not compelled), Mentally informed (not misinformed), Psychologically reasonable (not drugged or altered).
  • People can escape their own past decisions. (Dopamine rushes from Instagram Likes don’t trap users in trivial pursuits.)
  • People can make decisions. (Amazon displays products to provide choice, not to nudge choosing.)
  • People can experiment with new decisions: users have access to opportunities and possibilities (professional, romantic, cultural, intellectual) outside those already established by their personal data.

Dignity

  • Humans hold intrinsic value: they are ends in selves and not only means to other ends. People are not tools that can be simply used. People are not replaceable like commodities. Kant.
  • People treated as individually responsible for – and made responsible for – their choices. (Freedom from patronization, condescension, and pity.) Kant, Nietzsche.
    • The AI/human distinction remains clear. (Is the voice human, or a chatbot?)

Privacy

  • Control over access to your personally identifying information. Westin.
  • People can maintain multiple, independent personal information identities. (Work-life, family-life, romantic-life, can remain unmixed online.)
Values

Fairness

  • Equals treated equally and unequals treated proportionately unequally, within decision domain. Aristotle.
  • Process justifies the outcome, as opposed to outcome justified the process.
  • Equality applies to opportunities, not outcomes.
  • Equality about verbs (what you can do), not nouns (who/what you are).
  • Society is a collection of individuals, as opposed to individuals being aspects of a larger society.
  • Ideal treatment of individuals dictates subsequent society, as opposed to ideal vision of society dictating subsequent treatment of individuals.
  • Paradigmatic case (where most everyone wants fairness, not equity): the Olympics.
  • Everyone can advance, as opposed to no one left behind.

Equity/Solidarity

  • Greatest advantage to the least advantaged. (Max/Min distribution.) Rawls.
  • Outcome justifies the process, as opposed to process justifying the outcome.
  • Equality applies to outcomes, not opportunities.
  • Equality is about nouns (who you are), not verbs (what you can do).
  • Individuals are aspects of a larger society, as opposed to society being a collection of individuals.
  • Ideal vision of society dictates subsequent treatment of individuals, as opposed to the ideal treatment of individuals dictating subsequent society.
  • Paradigmatic case (where most everyone wants equity, not fairness): a family
  • No one left behind, as opposed to everyone can advance.

Social wellbeing/Sustainability

  • Greatest good to the greatest number. Utilitarianism, Bentham, Stuart Mill.
  • Good understood as good or happiness, and in reciprocal relation with pain and suffering.
  • Individual subordinated to the collective.
Values

Performance

  • Accuracy and efficiency in completing the assigned task.
  • End-in-self, not mediated value. (How well task is done, regardless of task's desirability or downstream effects in the world.)
    • Engineering evaluated like art, as intrinsic value, not in terms of utility.
  • Optimization in terms of sensitivity and selectivity, robustness, scalability.

Safety

  • Predictable, controllable, and aligned with human values.
  • Risks and harms minimized, even when confronted with unexpected changes, anomalies, and perturbations. Robust and resilient.
    • Tesla that functions well in San Francisco also does well in Trento, Italy.
    • Mental health conversational agent produces no shortcircuits across range of patients/vulnerabilities.

Explainability/Accountability

  • Explainability: Describe why/how output was produced in the languages of the various stakeholders
    • User: interpretability (''What do I need to change about myself to get a different result?'')
    • Developer: transparency (''What do I need to change in the Al to get a different result?'')
    • Lawyer: accountability (''Who can I credit/blame for Al performance?'')
  • The Delphic dilemma: Which is primary, making AI knowledge production better, or knowing why AI knows? What’s worth more, understanding or knowledge? (Knowing, or knowing why you know?)
    • If an AI picks stocks, predicts satisfying career choices, or detects cancer, but only if no one can understand how the machine generates knowledge, should it be used?
  • Accountability gap: Case where tool not explainable/impossible to attribute responsibility for action.
    • As a condition of the possibility of producing knowledge exclusively through correlation, AI may not be explainable.
  • Accountability gap dilemma: Which is primary, making AI better, or knowing who to blame, and why, when it fails?
  • Accountability ambiguity: A driverless car AI system refines its algorithms by imitating driving habits of the human owner (driving distance between cars, accelerating, breaking, turning radiuses). The car later crashes. Who is to blame?
  • Redress-by-design: Ability to attribute responsibility and compensate harm engineered from beginning.

Compared to others, our principles lean toward human freedom/libertarianism, and are more streamlined. Small differences.

Ethics Guidelines for Trustworthy AI
AI High Level Expert Group, European Commission
https://ec.europa.eu/digital-single-market/en/news/ethics-guidelines-trustworthy-ai

European Ethical Charter on the Use of Artificial Intelligence in Judicial Systems and their Environment
European Commission for the Efficiency of Justice
https://rm.coe.int/ethical-charter-en-for-publication-4-december-2018/16808f699c

Ethical and Societal Implications of Data and AI
Nuffield Foundation
https://www.nuffieldfoundation.org/sites/default/files/files/Ethical-and-Societal-Implications-of-Data-and-AI-report-Sheffield-Foundat.pdf

The Five Principles Key to Any Ethical Framework for AI
New Statesman, Luciano Floridi and Lord Clement-Jones
https://tech.newstatesman.com/policy/ai-ethics-framework

Postscript on Societies of Control
October 1997, Deleuze
/Library: Deleuze, Foucault, Discipline, Control.pdf

A Declaration of the Independence of Cyberspace
John Perry Barlow
https://www.eff.org/cyberspace-independence

 

 

Ethics Theories

S image

Deontological

Guiding value

Tradition

Rules for action

A set of moral directives seems to recur through historical times and places, and in diverse religious, social and political contexts. This endurability can be taken to legitimize the guidelines. While no single list perfectly contains the recurring imperatives, typically there are:

Duties to self:
• Preservation
• Develop my own talents
• Fidelity/Integrity (Be true to myself)

Duties to others:
• Honesty (Be true to others)
• Beneficence (Help others as reasonably possible)
• Reparation (Repair harm done to others)
• Gratitude

Advantages/Drawbacks

Because ethical legitimacy stands on widespread historical acceptance of the moral rules, the guidelines are familiar, commonly employed, and easily applied to experience. But, multiple duties may yield contradictory imperatives. For example, a student may have money to buy a new computer (develop own talents) or donate to a scholarship fund (beneficence), but not both. No formula has been discovered to reliably adjudicate these duty conflicts.

S image

Deontological

Guiding value

Religious faith

Rule for action

Follow the command of God (or Gods in the case of polytheism, as in ancient Greece).

Advantages/Drawbacks

Divine sanction fortifies confidence in moral regulation, but difficulties remain in decoding how the regulation should be applied on the human level, as exemplified by conflicting interpretations of religious texts, and by the story of Job in the Bible.

S image

Deontological

Guiding value

Equality

Rule for action (Aristotle version)

Treat people identically unless they differ in ways relevant to the situation. Differences between people that are relevant should yield proportionately unequal treatment. (Treat equals equally, and unequals unequally.)

Advantages/Drawbacks

Aristotelian fairness yields objectively correct responses to dilemmas. But, it can be difficult to define the “equal” and “unequal” in practice, especially in terms of what counts as relevant to a situation. For example, a five foot woman and a six foot man each pay the same price for an airplane ticket. Should they receive the same legroom?

Rule for action (John Rawls version)

Decide without regard for how your conclusion affects you personally. The theory can be presented as a thought experiment in which deciders know nothing about themselves (age, education, preferences, and so on) and after making a judgement, those qualities are assigned to them randomly. So, with respect to the airplane ticket and legroom, deciders must imagine that their height will be assigned by lottery after pronouncing their decision.

Advantages/Drawbacks

An effective strategy in some situations. For example, when sharing a cookie between two friends, one breaks it in half and the other chooses the side: the person breaking the cookie operates from behind the veil of ignorance in that they don’t know how they will be affected by their own portioning. But, in many situations it’s nearly impossible for deciders to blindfold themselves to their own reality within the decision being made.

S image

Deontological

Guiding values

Rationality, Dignity

Rule for action (Rationality version)

Actions must be universalizable, meaning that it is possible to rationally conceive of everyone taking the action all the time. Lying, for example, cannot be universalized because if everyone lied all the time, no one would take anything seriously, so no one could successfully lie. Attempting to lie therefore contradicts itself. Restated, lying cannot make sense because universalizing the practice doesn’t make everything false, instead, it creates a reality like an adventure movie which is neither honest nor dishonest: it’s not true, but it’s also not misleading, just entertaining.

Advantages/Drawbacks

Powerfully objective, but practically torturous: imagine never lying about anything, ever.

Rule for action (Dignity version)

Treat others as ends in themselves, and never only as means. Because others’ independent life projects must be respected, treating them as tools or instruments serving my own projects becomes inadmissible. The difference can be understood in the distinction between collaboration and exploitation: the first treats others as ends in themselves, the second treats them as tools for use.

Advantages/Drawbacks

The ideal of universal dignity as inherent to human being is inspiring in the abstract, but does a remorseless murderer deserve to be treated as dignified? More practically, if humans may not be treated as mere instruments, what does that mean for our interactions with cashiers?

S image

Deontological

Guiding value

Freedom

Rule for action

Do what you want, up to the point where you interfere with others doing the same. Libertarian models extend freedom expressions from our minds and bodies, to our possessions and the fruits of our labors. In every case, freedom means applying rules to yourself, and obeying them.

Advantages/Drawbacks

Freedom maximization empowers individual experience: we are liberated to choose our own identities and destinies. But, the theory does little to resolve conflicts between individuals or support collective wellbeing. Zoning laws, for example, conflict with libertarian thought.

S image

Consequentialist

Guiding value

Happiness

Rule for action

Bring the greatest good and happiness to the greatest number. Total wellbeing is calculated by summing the condition of every member of society. Then, those actions raising the happiness count – or diminishing overall suffering – are implemented. Happiness can be defined hedonically (Bentham, physical pleasures), or idealistically (Mill, intellectual pleasures). In both cases, the happiness calculation must account for everyone, as far into the future as the effects of an action may be reasonably projected.

Advantages/Drawbacks

Overall wellbeing and the collective welfare is attractive in the abstract, but balances against injustices to flesh and blood individuals: if a fatal disease can be cured with a lethal experiment on a human, and there are no volunteers, a pure utilitarian will coerce participation. Another drawback is the difficulty in accurately calculating happiness in a world of diverse people with unpredictable futures.

S image

Consequentialist

Guiding value

Happiness

Rule for action

Bring the greatest good and happiness to the greatest number, not including the actor. Total wellbeing is calculated by summing the condition of every member of society except the person doing the calculating. Then, those actions raising the happiness count – or diminishing overall suffering – are implemented. Happiness can be defined hedonically (Bentham, physical pleasures), or idealistically (Mill, intellectual pleasures). In both cases, the happiness calculation must account for everyone except the actor, and the calculating must stretch as far into the future as the action’s effects may be reasonably projected.

Advantages/Drawbacks

Selflessly seeking collective wellbeing sounds noble, but is altruism based on generosity, or is it disguised self-abnegation?

S image

Consequentialist

Guiding value

Happiness

Rule for action

Bring the greatest good and happiness to me. Happiness can be defined hedonically (Bentham, physical pleasures), or idealistically (Mill, intellectual pleasures). Some egoists view the ethics as an inescapable psychological reality: we are all out for ourselves whether we admit it or not. Others support egoism as a rational choice, especially those promoting Enlightened Egoism, the view that acting to benefit others is desirable as an efficient strategy for self-service. More, the best way to bring happiness to others may be to seek it for oneself (Adam Smith, invisible hand).

Advantages/Drawbacks

No one knows my own happiness better than I do, so it makes sense that I hold the responsibility to seek it. Also, if the invisible hand idea is persuasive, then enlightened egoism becomes preferable to utilitarianism and altruism by default. But, egoism requires that others, even those closest to us, be categorized as unworthy of independent moral consideration.

S image

Virtue

Guiding value

Good living (Eudaimonia)

Rule for action

As opposed to deontological and consequentialist theories which attempt to form good rules for action, virtue ethics attempts to form good people, and then trust that they will act civically in a complex world. Virtue is a skill, one that is acquired intellectually through study, and also practically as part of youthful development in social institutions: families, schools, churches, the military, workplaces, sports teams, civic associations. As an example of virtue happening, a college student may attend an ethics lecture in the afternoon, and in the evening apply the lessons while participating in a water polo competition where the virtue of winning with humility (or losing with dignity) is practiced.

Advantages/Drawbacks

Because virtue is a skill, the attainment of mastery provides satisfaction, meaning virtue is its own reward: doing good feels good, and together they define a good life. However, exactly what counts as being virtuous is hard to define since different societies teach their youth different lessons and embody distinct practices for managing crime and punishment, vows of marriage and family responsibilities, the treatment of the vanquished in war and sporting competitions, and so on.

S image

Post-Nietzschean (Nietzsche/Heidegger)

Guiding value

Authenticity

Rule for action

As opposed to the traditional ethical obligation for individuals to aspire to an ideal identity as defined by their society, the ethics of authenticity asks you to be true only to yourself, whoever you may be. The precedent requirement is to determine who, exactly, you are. Nietzsche proposed the Eternal Return thought experiment, Heidegger proposed anxiety in the face of death. In both cases, the result is an understanding one’s unique life projects as distinct from broader social expectations. The subsequent ethical imperative is to engage those projects.

Advantages/Drawbacks

In a world without objective right and wrong, being true to myself provides a direction and use for my freedom. When the authentic person is an artist, the theory works well, but when the authentic person is a natural born murderer, the theory is less felicitous.

S image

Post-Nietzschean

Guiding value

Nativity

Rule for action

Traditionally, ethicists have worked to escape the idiosyncrasies of particular times and places by developing theories sufficiently abstract to apply universally. Culturalism reverses the tradition by embracing the idiosyncrasies: a community’s native beliefs are accepted as their legitimate moral rules, and the task of ethics is to learn the local practices, customs, and traditions, and then fit into them.

Advantages/Drawbacks

Respect for distinct cultures and traditions is maximized, but hope for ethical progress recedes because respecting another culture’s moral rules goes equally whether those rules seem noble, or barbaric.

S image

Post-Nietzschean (Habermas)

Guiding value

Consensus

Rule for action

Gather those involved in a conflict and discuss until reaching a shared resolution. The discourse must be rational and peaceful: participants comprehend their own agreements, and arrive without coercion. There is a partial analogy to American courthouse jury decisions here in that agreement is by informed consent, and the fact of agreement is the decision’s judicial/ethical legitimacy.

Advantages/Drawbacks

Provides a broad range of initially possible solutions since everything is on the table for discussion. But, everything on the table means a lot of talking since every conflict must be addressed and resolved from scratch.

S image

Post-Nietzschean (Gilligan)

Guiding value

Care

Rule for action

As opposed to concentrating on individuals, care ethics focuses on the links uniting people in their social networks. The aim is to strengthen the web of bonds, especially with those who are nearest. Families are a commonly cited example. For instance, a relative suffering drug addiction may receive a disproportionately large share of resources and concern. Or, if the addict becomes dangerously toxic, links to the family may be severed. In both scenarios, fortifying the web of familial care is the guiding concern, not any particular member.

Advantages/Drawbacks

Fortifying our intimate social networks conforms to intuitive feelings: many of us would rescue a sister before a stranger if only one could be saved. But, the theory can lead to tribalism, a mafia-family approach to civil co-existence.

S image

Post-Nietzschean (Deleuze)

Guiding value

Originality

Rule for action

The traditional ethical split between wrong and right is replaced by stagnation and creation. Creativity as an ethics repurposes customary elements of experience for original uses. A common example is slang: redirecting a language’s standard words for divergent meanings. Another example is the reorienting of web platforms, including the exploitation of LinkedIn as a dating site. As an ethics, creativity works within its native reality instead of coming from outside, its twists the elements of experience away from orthodox uses as opposed to destroying them, and it escapes conventions as opposed to overthrowing them.

Advantages/Drawbacks

Originality as the highest value can be individually invigorating. But, if the driving reason we innovate is to go on and create something else, the interminability is daunting, as in the endless “Yes, and…” of improvisation theater.