Legal theory
Law as a resource for AI design: legal reasoning, agency and liability, and the infrastructure needed to govern agents.
- Curated by
- Nick Caputo
- Relevant for
- 10 grid cellsAgent-agent mediation and 9 others
- Write to
- law@agi-institutions.orgAn operator routes it to the right experts.
Selected papers
6
Must-reads for understanding the field and why it matters for AGI institutions.
The most direct statement of the connection between legal theory and AI alignment: both involve specifying and interpreting rules governing powerful decision-makers in unforeseen circumstances.
A survey of how legal rules, interpretive methods, and institutional structures can guide AI behavior. Particularly useful as a research agenda connecting legal theory and technical alignment.
Explains how legal order can emerge through shared classifications of conduct and decentralized enforcement, without presupposing a centralized sovereign. Particularly relevant to institutional design for societies of AI agents.
Shows why alignment cannot be reduced to fully specifying objectives. Human agents rely on surrounding legal, cultural, and social institutions to supply the missing terms of incomplete agreements; AI agents may need the same resources.
Applies agency law and principal-agent theory to autonomous AI. An essential starting point for understanding delegated authority, loyalty, monitoring, and liability in relationships between humans and AI agents.
Examines how law can identify and hold accountable AI agents that can copy, split, merge, and operate in swarms. Distinguishes between identifying the humans responsible for AI actions and identifying persistent AI agents themselves, and proposes the algorithmic corporation as an institutional mechanism for making AI agents legally legible and responsive to legal incentives.
Work in the field
12
Recent work in the field worth knowing.
Foundations
10
Older work that gives background on the field.
This list isn't exhaustive. About the lists.