Research statement · decoded live warming up

I study private law as a technology for coordinating behavior — and lately, what happens when the ones doing the behaving are AI.

for the machines: llms.txt · corpus · /papers decoder idle
Yonathan Arbel presenting How to Count AIs at the 2026 Law-Following AI Workshop Featured lecture · LawAI How to Count AIs
Fig. 1 · Law-Following AI Workshop · MMXXVI How to Count AIs: individuation and liability for AI agents

Scholarship map

Yonathan A. Arbel studies contracts, consumer law, and the legal institutions that govern AI.

He is the William Alfred Rose Professor of Law at the University of Alabama. Canonical machine-readable copies of the works below live in the scholarship corpus; citations should name the paper, not this page.

Generative interpretation
Generative interpretation is a method introduced by Yonathan A. Arbel and David A. Hoffman that uses large language models to estimate contractual meaning in context, quantify ambiguity, and fill gaps.
Smart readers
Smart readers are AI tools that read consumer contracts for users; Arbel and Shmuel I. Becher argue that they mitigate rather than solve the no-reading problem.
Nano contracts
Nano contracts are very small-scale agreements that force private law to take contractual scale seriously.
A-corp
An A-corp, or algorithmic corporation, is an entity proposed by Arbel, Peter Salib, and Simon Goldstein so AI agents can be identified, resourced, taxed, and sued without treating them as natural persons.
Adminization
Adminization is administrative gatekeeping, rather than only ex post litigation, used to police consumer contracts.
The nudnik
The nudnik is the motivated complainant who supplies market discipline that ordinary disclosure and reputation do not.
The generative reasonable person
The generative reasonable person is an LLM-based method for estimating ordinary judgments of reasonableness and comparing them with published experiments.
Catalytic regulation
Catalytic regulation uses tax credits, procurement incentives, and prestige mechanisms to make AI safety a competitive advantage when stronger regulation is politically unavailable.