As large language models (LLMs) increasingly mediate ethically sensitive decisions, understanding their moral reasoning processes becomes imperative. This study presents a comprehensive empirical evaluation of 14 leading LLMs, both reasoning enabled and general purpose, across 27 diverse trolley problem scenarios, framed by ten moral philosophies, including utilitarianism, deontology, and altruism. Using a factorial prompting protocol, we elicited 3,780 binary decisions and natural language justifications, enabling analysis along axes of decisional assertiveness, explanation answer consistency, public moral alignment, and sensitivity to ethically irrelevant cues. Our findings reveal significant variability across ethical frames and model types: reasoning enhanced models demonstrate greater decisiveness and structured justifications, yet do not always align better with human consensus. Notably, "sweet zones" emerge in altruistic, fairness, and virtue ethics framings, where models achieve a balance of high intervention rates, low explanation conflict, and minimal divergence from aggregated human judgments. However, models diverge under frames emphasizing kinship, legality, or self interest, often producing ethically controversial outcomes. These patterns suggest that moral prompting is not only a behavioral modifier but also a diagnostic tool for uncovering latent alignment philosophies across providers. We advocate for moral reasoning to become a primary axis in LLM alignment, calling for standardized benchmarks that evaluate not just what LLMs decide, but how and why.

In 1967, British philosopher Philippa Foot unwittingly created one of the internet's most regurgitated memes. A runaway train is hurtling towards five people tied to the tracks. You can pull a lever to divert the train to a different track to which only one person is tied. Do you intervene to kill the one and spare the five? What if one of the tracks twisted into a really cool loop-the-loop?

This study examines the ethical reasoning of six prominent generative large language models: OpenAI GPT-4o, Meta LLaMA 3.1, Perplexity, Anthropic Claude 3.5 Sonnet, Google Gemini, and Mistral 7B. The research explores how these models articulate and apply ethical logic, particularly in response to moral dilemmas such as the Trolley Problem, and Heinz Dilemma. Departing from traditional alignment studies, the study adopts an explainability-transparency framework, prompting models to explain their ethical reasoning. This approach is analyzed through three established ethical typologies: the consequentialist-deontological analytic, Moral Foundations Theory, and the Kohlberg Stages of Moral Development Model. Findings reveal that LLMs exhibit largely convergent ethical logic, marked by a rationalist, consequentialist emphasis, with decisions often prioritizing harm minimization and fairness. Despite similarities in pre-training and model architecture, a mixture of nuanced and significant differences in ethical reasoning emerge across models, reflecting variations in fine-tuning and post-training processes. The models consistently display erudition, caution, and self-awareness, presenting ethical reasoning akin to a graduate-level discourse in moral philosophy. In striking uniformity these systems all describe their ethical reasoning as more sophisticated than what is characteristic of typical human moral logic.

The Trolley Problem has inspired scores of psychology experiments, including MIT's Moral Machine,1 an online survey where people had to decide what a self-driving car should do in case of an impending accident. Participants were given a series of pairs of scenarios, presented as map-like diagrams, with various numbers and types of pedestrians and passengers. For each pair of scenarios, they had to choose between options such as driving ahead and killing pedestrians, or veering into an obstacle and killing passengers. Based on 40 million responses from more than 200 countries, they found general preferences, such as sparing humans over animals. They also found differences between cultures. People from countries with collectivistic cultures prefer sparing lives of older people instead of the lives of younger people.

You want me to choose whether we have red or white wine? First, let me tell you about being abducted by aliens. I was standing on Westminster Bridge in London, and Big Ben had just chimed the hour. Next moment, I am on the bridge of a starship, face-to-face with the pointy-eared alien from that '60s sci-fi show. "Either this is a dream, or something's interfering with my mind. "We thought this would make the transition easier for you." The only realistic way to travel across the galaxy is as an artificial intelligence. Our ship is crewed by AIs." Why?" Somehow, I'd expected first contact with aliens to be more profound, but then I didn't do it every day. "Even your primitive designs have benefited from interaction between AIs--it's the best way to enable machine learning.

An expert in computing responds to Tara Isabella Burton's "I Know Thy Works." In 2018 researchers from the Massachusetts Institute of Technology Media Lab, Harvard University, the University of British Columbia, and Université Toulouse Capitole shared the results of one of the largest moral experiments conducted to date. They recorded 40 million ethical decisions from millions of people across 233 countries. The experiment's "Moral Machine" posed to users variations of the classic trolley problem, imagining instead the trolley as a self-driving car. Should the car swerve and collide with jaywalking pedestrians or maintain its current trajectory, which would yield inevitable doom for the passengers inside?

Researchers at TH Ingolstadt and the University of Southern Denmark have studied the effects of AI opinions on humans. Their study shows that machine-generated moral perspectives can influence people, even when they know the perspective comes from a machine. In their two-step experiment, the researchers first asked ChatGPT to find solutions to different variants of the trolley problem: Is it right to sacrifice the life of one person to save the lives of five others? The researchers received different advice from ChatGPT. Sometimes the machine argued for human sacrifice, sometimes against.

Chris Dixon, who invested early in companies ranging from Warby Parker to Kickstarter, once wrote that the next big thing always starts out looking like a toy. That's certainly true of artificial intelligence, which started out playing games like chess, go and playing humans on the game show Jeopardy! Yet today, AI has become so pervasive we often don't even recognize it anymore. Besides enabling us to speak to our phones and get answers back, intelligent algorithms are often working in the background, providing things like predictive maintenance for machinery and automating basic software tasks. As the technology becomes more powerful, it's also forcing us to ask some uncomfortable questions that were once more in the realm of science fiction or late-night dorm room discussions.

OpenAI's conversational language model ChatGPT has a lot to say, but is likely to lead you astray if you ask it for moral guidance. Introduced in November, ChatGPT is the latest of several recently released AI models eliciting interest and concern about the commercial and social implications of mechanized content recombination and regurgitation. These include DALL-E, Stable Diffusion, Codex, and GPT-3. While DALL-E and Stable Diffusion have raised eyebrows, funding, and litigation by ingesting art without permission and reconstituting strangely familiar, sometimes evocative imagery on demand, ChatGPT has been answering query prompts with passable coherence. That being the standard for public discourse, pundits have been sufficiently wowed that they foresee some future iteration of an AI-informed chatbot challenging the supremacy of Google Search and do all sorts of other once primarily human labor, such as writing inaccurate financial news or increasing the supply of insecure code.

In my book'Rise of the Moral Machine: Exploring Virtue Through a Robot's Eyes', I include a short fictional story about a couple (Mr and Mrs Morales) who are in the process of purchasing their first autonomous vehicle. Having chosen the model, the colour and the trim of the car, the last set of choices they are required to make concern the vehicle's'ethical alignment': i.e. the alignment of the vehicle's autonomous decisions on how it should drive with the Morales' social and ethical preferences. Without giving too much of the story away, the Morales' are presented with a series of situations each of which requires the autonomous vehicle to make a moral decision. These decisions are presented in terms of choices of who should be the casualties of an unavoidable collision, such as "should the vehicle run over the pensioner on the pedestrian crossing, or the child on the pavement?" (Figure 1).