Record STMT-WD6E44BAspirational

Understanding a Robot’s Guiding Ethical Principles via Automatically Generated Explanations

White paper State lawGlobaladopted

This research paper examines how contrastive and non-contrastive explanations can help humans understand the ethical principles underlying a robot's action plans. The authors present a formal framework that maps ethical principles like deontology, utilitarianism, and do-no-harm into automated planning models and constraint operators to generate these explanations. Results from a user study indicate that both explanation types help people understand moral reasoning in robots, though with varying nuances.

Summary and content fields are drawn from the document text and are pending editorial review.

Who issued it

arXivAcademic

Publisher

Legal forceAspirational

Under whose authorityState law

Judged fromfull text

Who it governs

Researchers and academicsAI developers and providers

AI researchers and developers interested in explainable automated planning and machine ethics should read this to understand how to generate and evaluate moral explanations for autonomous systems.

Sectors

Research and science · artificial intelligence and robotics research

What the document contains

What's covered

The paper covers a technical approach and formalisms for generating non-contrastive and contrastive explanations for the ethics of robot action plans using AI planning models. It also describes an iterative dialogue framework for users to explore alternative plans and an empirical user study evaluating the effectiveness of these explanations.

What's missing

The text does not provide commercial software implementation guidelines, legislative enforcement mechanisms, or deployment frameworks for real-world physical robotics outside of abstract planning models.

Key takeaways

Automated planning systems can be constrained to adhere to specific ethical principles such as deontology, utilitarianism, and do-no-harm.
Contrastive explanations allow users to ask 'Why A rather than B?' questions regarding a robot's ethical choices.
A user study demonstrated that providing explanations significantly helps humans build a mental model of a robot's moral reasoning.
The empirical evaluation showed that participants successfully transferred their understanding of robot moral reasoning from the trolley problem to the bridge dilemma.

Principles it states

(Act-)Deontology: plans must not contain any inherently bad actions.Utilitarianism: plans must lead to one of the best reachable world states.Do-No-Harm principle: plans must not be the cause of any bad outcomes.

Outline

· I. Introduction
· II. Related Work
· III. Background
· IV. Explaining Robot Action Plans
· V. Empirical Evaluation of the Generated Explanations
· VI. Conclusions

Issued20 June 2022

Length7,272 words · 8 pages

FormatNot recorded

5/5

Openness about how AI systems work and are used

This paper presents a technical approach to generating non-contrastive and contrastive explanations in the context of ethical AI task planning.

Section I

screening score · 2026-10-24

Enforcement

0/5

Consequences and compliance mechanisms with real force

N/A

screening score · 2026-10-24

Impact Assessment

0/5

Required assessment of an AI system's risks/impacts before or during use

N/A

screening score · 2026-10-24

Topics

Physical AI Governancesupporting

explainable_ai robot_ethics automated_planning moral_reasoning contrastive_explanations