Can AI Teach Values? Why Machines Struggle with Morality

Jun 16

Artificial intelligence can beat chess champions, drive cars, and write essays. But there’s one area where it consistently struggles: morality. Values like fairness, compassion, and justice are deeply human concepts shaped by culture, history, and lived experience. The question is, can AI ever truly understand them—or is morality something machines will never master?

The Problem of Programming Values

At its core, AI makes predictions based on data. It doesn’t “understand” right or wrong; it only follows patterns. If we want AI to act ethically, humans have to program rules. But whose rules? What one culture considers fair may look different in another. For example, should a healthcare AI prioritize the youngest patients because they have more years to live, or the sickest patients because they need help most urgently?

When AI Faces Ethical Dilemmas

Self-driving cars highlight the challenge. Imagine a scenario where a crash is unavoidable: should the car protect its passengers at all costs, or minimize harm to pedestrians? Humans debate such questions endlessly. Expecting a machine to resolve them fairly is even more difficult.

The Risk of Bias

AI often inherits the values, both good and bad, of the data it is trained on. If historical hiring data favors men, the AI may “learn” that bias. If law enforcement data reflects over-policing in certain communities, the AI may reproduce those injustices. In other words, morality in AI isn’t neutral; it reflects the choices and prejudices of the people who built it.

Can AI Learn Morality?

Some researchers are exploring ways to teach AI systems ethical reasoning, using philosophical principles or crowdsourced opinions. But morality isn’t just logic, it’s emotion, empathy, and lived context. Machines may get better at mimicking ethical reasoning, but whether they can ever feel values is another matter entirely.

Closing Thoughts: The Human Role in Teaching Machines

AI will never fully grasp morality because morality is not just a set of rules, it’s a shared human experience. Machines may help us act on our values more consistently, but they cannot define them. For teenagers, this means recognizing the responsibility of shaping technology’s direction. For elders, it’s a reminder that wisdom built from experience will always be needed to guide new tools. AI may calculate outcomes, but humans must decide what is right.

Tanya Patel

Tanya Patel is a senior at The Pingry School with a strong academic focus on economics, business, finance, and accounting. She is the founder and president of Farming for GRACE, a student-led initiative that grows and donates culturally relevant produce. She also mentors children and provides health app support to elders at her temple and coaches youth soccer. Across all of her endeavors, Tanya is motivated by one throughline: ensuring systems—whether in food, technology, healthcare, or community—are built with equity, dignity, and inclusion at their core