Ethical Decision Making in Autonomous Vehicle Systems

Ethical decision making in autonomous vehicle systems addresses how self-driving cars should be programmed to act in situations where harm is unavoidable, how they balance competing obligations to passengers, pedestrians and other road users, and how societal values are encoded into algorithms. Central concerns include the design of decision rules for emergencies, assignment of responsibility in the event of collisions, and mitigation of human biases in moral judgments. Research spans philosophical foundations—drawing on deontological and consequentialist theories—to applied frameworks for trajectory planning and control. Ethical frameworks are increasingly integrated into technical architectures, ensuring that safety objectives align with public expectations and regulatory standards. The global significance of this work lies in reducing road fatalities, fostering public trust in automated mobility, and informing policy on liability, data governance and equitable access to emerging transport modes.

Research from Nature Portfolio

Studies have demonstrated that human moral biases profoundly affect the design of universal decision rules for autonomous vehicles. Under intuitive decision-making, individuals tend towards deontological choices, favouring preservation of pedestrians over passengers, whereas a more deliberative mode shifts preferences towards utilitarian outcomes. Moreover, the perspective of the decision-maker influences moral choices: when identifying with the passenger, participants prioritise self-protection, but when viewing the scenario impartially they support actions that minimise total harm. These findings underscore the challenge of reconciling individual self-interest with collective welfare and highlight the need for algorithms that account for systematic biases in moral judgment without exacerbating social dilemmas.

Ethical Decision Making in Autonomous Vehicle Systems publication trend

The graph below shows the total number of articles in ethical decision making in autonomous vehicle systems across all publications each year (not limited to Nature Index journals).

Technical terms

Deontological ethics: An ethical framework that emphasises adherence to rules or duties regardless of consequences.

Consequentialism: An ethical approach in which the rightness of an action is judged solely by its outcomes, often framed in terms of utility.

Utilitarianism: A form of consequentialism that seeks to maximise overall welfare or minimise total harm.

Trolley problem: A thought experiment presenting a moral dilemma in which action must be chosen between harming one individual to save many or refraining from intervention.

Moral dilemma: A situation in which there is a conflict between two or more ethical principles, and any choice violates at least one principle.

Algorithmic bias: Systematic deviation in algorithmic decisions reflecting skewed data or embedded human values, potentially leading to unfair outcomes.

References

1. When own interest stands against the “greater good” – Decision randomization in ethical dilemmas of autonomous systems that involve their user’s self-interest. Computers in Human Behavior Artificial Humans (2024).
2. The Ethics of Accident-Algorithms for Self-Driving Cars: an Applied Trolley Problem?. Ethical Theory and Moral Practice (2016).
3. The Future of Transportation: Ethical, Legal, Social and Economic Impacts of Self-driving Vehicles in the Year 2025. Science and Engineering Ethics (2019).
4. Incorporating Ethical Considerations Into Automated Vehicle Control. IEEE Transactions on Intelligent Transportation Systems (2017).
5. Human decision-making biases in the moral dilemmas of autonomous vehicles. Scientific Reports (2019).
6. Moral Judgements on the Actions of Self-Driving Cars and Human Drivers in Dilemma Situations From Different Perspectives. Frontiers in Psychology (2019).