arXiv:2607.00002v1 Announce Type: new Abstract: Moral cognition has traditionally been modeled as adherence to fixed ethical theories--deontology, consequentialism, virtue ethics--implemented as static rules or value functions. We propose Bounded Morality, a formal framework for analyzing the computational demands of moral problems faced by finite agents. Extending Herbert Simon's notion of bounded rationality, we formalize moral situations along two orthogonal dimensions: moral breadth, the scope of entities treated as morally relevant, and moral depth, the inferential integration required to
The rapid advancement and deployment of AI systems necessitate a renewed focus on formalizing moral cognition to ensure their responsible operation and integration into society.
This framework offers a critical step towards building more robust, transparent, and ethically aligned AI, moving beyond static rules to address the complexities of real-world moral decision-making.
The definition of moral computation shifts from rigid ethical theories to a more dynamic model accounting for computational limits and the scope of moral considerations in AI systems.
- · AI ethics researchers
- · Developers of autonomous AI systems
- · Regulatory bodies
- · Philosophers of AI
- · Developers of 'black box' AI
- · AI systems lacking explicit ethical frameworks
AI development incorporates 'bounded morality' as a design principle for ethical decision-making.
Improved public trust and acceptance of advanced autonomous AI systems due to transparent moral reasoning.
The emergence of new AI system architectures explicitly designed to quantify and manage 'moral breadth' and 'moral depth'.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI