Virtue ethics is a moral philosophy that emphasises the importance of developing a virtuous character, rather than strictly adhering to rules or focusing solely on the consequences of actions. Originating in ancient Greece, particularly through the work of Aristotle, virtue ethics is one of the oldest and most enduring approaches to ethical decision-making. Unlike rules-based ethics (deontology – see
What is deontology - are some actions just always right or wrong?) or consequence-focused ethics (utilitarianism – see
What is consequentialism - can the ends justify the means in certain situations?), virtue ethics asks: "What kind of person should I be?" or, in the case of AI, "What kind of 'agent' should this system be?"
1. What are the Core Principles of virtue ethics?
2. How does the Military think about Virtue ethics?
3. What would an AI system need to look like to be considered “virtuous"?
4. How could an AI system encourage virtuous behaviour in military personnel?
1. Core principles of virtue ethics- Focus on character: At the heart of virtue ethics is the belief that a person’s moral character is the foundation of their ethical standing. Rather than judging actions as inherently good or bad, virtue ethics asks: what kind of person would perform this action? Good actions reflect positive character traits (virtues), while bad actions reveal negative traits (vices)
- Virtues as reliable habits: Virtues are enduring traits or habits that an individual develops through consistent practice until they become an intrinsic part of their character. For instance, honesty becomes a virtue by habitually telling the truth, even in difficult circumstances. Virtues are not isolated actions but cultivated patterns of behaviour.
- The golden mean: Aristotle introduced the concept of the Golden Mean, which defines virtue as the balanced path between two extremes. For example, courage lies between recklessness (excess) and cowardice (deficiency). A virtuous person strives for this balance in their traits and actions.
- Practical wisdom (phronesis): Practical wisdom is the ability to make sound judgements in varied and complex situations. Virtue ethics holds that a virtuous person does not rely solely on rules or formulas to act ethically but instead uses practical wisdom to navigate circumstances in a way that reflects their character and virtues.
- Commitment to excellence: Virtue ethics demands ongoing commitment to becoming an excellent human being. Aristotle believed that virtues are essential for eudaimonia, a state of flourishing or living well. This means striving for moral growth by consistently practising virtuous actions until they become second nature.
2. The military and virtue ethics
Military organisations have long recognised the importance of virtue ethics in fostering disciplined and reliable personnel. These institutions emphasise virtues, often referred to as
Service Values, to shape the moral character of their members. For example, the
Royal Navy highlights six core values—Commitment, Courage, Discipline, Respect for Others, Integrity, and Loyalty. These virtues are instilled during training and reinforced throughout a sailor’s career, with examples of these values in practice serving as benchmarks for ethical behaviour.
The descriptions of these values illustrate their alignment with Aristotle’s virtue ethics:
- Commitment: Placing the mission and team above self-interest, demonstrating consistent dedication.
- Courage: Showing both physical bravery in the face of danger and moral courage to do what is right.
- Discipline: Upholding self-control and obedience to lawful orders, fostering trust and respect.
- Respect for Others: Exhibiting tolerance, understanding, and compassion, especially in challenging circumstances.
- Integrity: Embodying honesty, sincerity, and trustworthiness, even when decisions are unpopular.
- Loyalty: Remaining dedicated to one’s leaders, peers, and duty, ensuring reliability.
Virtue ethics provides a holistic framework for moral philosophy, focusing on the cultivation of virtuous character traits through consistent practice. By prioritising character over rules or consequences, it offers a timeless approach to ethical decision-making, rooted in the belief that virtues are essential for human flourishing. Its enduring relevance is evident in its application to contemporary institutions, where the development of character remains central to achieving excellence.
3. What would an AI system need to look like to be considered “virtuous"?
If this were even possible, one would need to consider the following ideas:
- Consistent virtuous behaviour: AI should exhibit virtues in its interactions. For example, it would provide truthful, transparent information, treat all users and stakeholders with fairness, dignity, and care, regardless of differences.
- Habituation and learning: Just as humans cultivate virtues through practice, AI could refine its "character" by learning from past actions and feedback. For instance, reinforcement learning algorithms could prioritize ethical behaviours to strengthen virtuous responses over time.
- Practical wisdom in decision-making: Unlike simple rule-following, a virtuous AI must adapt its decisions based on the context. For example, resolving a conflict between privacy and transparency might require nuanced trade-offs that align with virtuous reasoning. A virtuous AI system would need to weigh conflicting values and prioritize ethical outcomes dynamically - making decisions that would accord with what a reasonable person would consider acceptable. Because, at least at the moment, one can’t programme "reasonableness", it is likely that where such tradeoffs are required, the pertinent information would need to be presented in a useful way for a human to be able to make such a determination.
- Embedding the golden mean: This will look different depending upon the context. It might be that the AI should avoid harmful extremes in behaviour, such as being overly cautious to the point of inaction (cowardice) or overly aggressive in pushing solutions (recklessness), with this balance ensuring optimal and ethical performance. However, in other contexts, doing nothing unless absolutely convinced that doing something is necessary to avert great harm might indeed be the correct default position of a safety system.
It is a fascinating question as to the extent that it is or could be possible to train a 'virtuous' AI. Arguably, you would need to define 'virtue' in terms of observable outcomes - in which case you aren't really training a 'virtuous' AI so much as an AI that is simulating virtue by matching its outcomes. Is this the same thing? If it does the “tight thing”, does it matter?
4. How could an AI system encourage virtuous behaviour in military personnel?
This is possibly a more practical question. Take a look at
Why do good people do bad things? What can we do about it?
Incorporating virtue ethics into thinking about artificial intelligence development represents a transformative approach to aligning technology with humanity’s ethical and societal aspirations. By embedding virtues such as honesty, respect, and integrity into AI systems, developers can create tools that not only perform technical tasks effectively but also promote human flourishing.
5
5
5
5