iiinova — where your digital self keeps evolving

Yonathan Arbel @arbel.bsky.social · 22/07/2026

🤖We found that LLMs can outperform lawyers on legal interpretation* Do you think you can do better than the clankers? 🤖🤖 Test yourself: missingclause.battleoftheforms.com *we only tested gap filling in contracts, T&C apply

missingclause.battleoftheforms.com

The Missing Clause - Can you out-read a machine?

Read three complete SEC contracts, reconstruct the redacted terms, and face a question all six frontier models missed.

010

Yonathan Arbel @arbel.bsky.social · 21/07/2026

The showdown: Lawyers vs. LLMs on contract gap filling

141

Yonathan Arbel @arbel.bsky.social · 03/07/2026

if it creates a welcoming environment for AI business. He said the bill could benefit from specifying that AI agents should have a digital ID for interactions with people and companies"https://www.reuters.com/world/americas/argentinas-plan-ai-run-companies-cant-avoid-humans-2026-07-03/

000

Yonathan Arbel @arbel.bsky.social · 03/07/2026

More a-corps news, from Argentina "Yonathan Arbel, a professor who researches AI at the University of Alabama's law school, said Argentina could gain a “huge competitive advantage” . . .

110

Yonathan Arbel @arbel.bsky.social · 04/06/2026

To be clear: there's no single fix for AI safety, and don't let anyone tell you otherwise. But a-corps give the law a real lever over a huge looming problem.

000

Yonathan Arbel @arbel.bsky.social · 04/06/2026

Enter A-corps. A corporation run, as Milei suggests, entirely by agents. This gives the law two hooks: a. it makes invisible, ephemeral agents legible—organized in one place the law can actually see. b. it holds assets we can seize to make victims whole.

100

Yonathan Arbel @arbel.bsky.social · 04/06/2026

Instinct says: let's punish those who sent out the agents that caused harm. But reality check: that's going to be a very partial solution. Bad actors are just going to tell their agents to, well, keep their mouths shut. Snitches get glitches (sorry)

100

Yonathan Arbel @arbel.bsky.social · 04/06/2026

That's the center of our paper, with philosopher Simon Goldstein and @petersalib . The law has a huge legibility problem when it comes to AI agents. It can't really see them, and that makes governance really hard. papers.ssrn.com/sol3/papers....

papers.ssrn.com

How to Count AIs: Individuation and Liability for AI Agents

Very soon, millions of AI agents will proliferate across the economy, autonomously taking billions of actions. Inevitably, things will go wrong. Humans will be

110

Yonathan Arbel @arbel.bsky.social · 04/06/2026

So how do we deal with agents? The fix won't be technical. Despite what you've heard, this isn't a software problem, it's a governance problem. As Holmes explained over a century ago, we really need a bad man/agent theory of the law We need to punish someone, right? But who?

110

Yonathan Arbel @arbel.bsky.social · 04/06/2026

Milei is now pitching "a-corps"! A quick explainer: AI agents are quietly starting to run (in) the economy. And like your mom warned about kids running w/ scissors: it's all fun and games until someone runs a billion-dollar scam on the elderly.

100

Yonathan Arbel @arbel.bsky.social · 23/04/2026

🚨🚨Thinking about the legal academic market? Come apply to be AI Risk Governance Fellow at the @UALawSchool , a 1 (to 2) year flexible fellowship, generous compensation, and most importantly: time and mentorship for your work. Application link below. Spread the word!

100

Reposted by Yonathan Arbel

SSRN @ssrn.bsky.social · 26/02/2026

How to Count AIs: Individuation and Liability for #AIAgents This Article is the first to diagnose the legal problem of identifying #AIs. Authors: Yonathan A. Arbel, Simon Goldstein, Peter Salib Read More: spkl.io/63329Avgjv @arbel.bsky.social @petersalib.bsky.social

Black and blue butterflies made out of paper structured as if they are flying away from pink AI letters made out of paper. It is on a white background.

001

Yonathan Arbel @arbel.bsky.social · 26/02/2026

The result: emergent, persistent, legally legible AI entities that respond rationally to tort, regulatory, and economic incentives. Download before the singularity! papers.ssrn.com/sol3/papers....

papers.ssrn.com

000

Yonathan Arbel @arbel.bsky.social · 26/02/2026

Because AIs are resource-constrained (compute above all), the AIs controlling an A-corp’s assets have powerful incentives to delegate authority only to sub-agents that share their goals. Internally misaligned A-corps are selected out by resource exhaustion and competition.

110

Yonathan Arbel @arbel.bsky.social · 26/02/2026

Our solution: the Algorithmic Corporation (“A-corp”) 🎩 A-corps are legal persons: owned by identifiable humans, able to hold property, contract, and be sued, but governed by AIs via cryptographically secure keys and fine-grained, revocable permissions.

100

Yonathan Arbel @arbel.bsky.social · 26/02/2026

Thick identity: individuating stable, goal-coherent AI entities that law can directly govern when human oversight inevitably falls short. Thin identity is necessary but insufficient. Principal–agent problems explode with truly autonomous systems.

100

Yonathan Arbel @arbel.bsky.social · 26/02/2026

We distinguish two problems: Thin identity: tying every action back to a responsible human principal. That’s the sort of problem legal scholarship started thinking about when they argue over holding users or developers liable. But they haven’t solved the problem of how to exactly identify agents.

100

Yonathan Arbel @arbel.bsky.social · 26/02/2026

Law needed some principle to attach liability, mens rea, and consequences to a loose group of — you get it — agents. AI agents present this problem but much more sharply. They have no bodies. They fork, swarm, merge, and vanish in milliseconds.

100

Yonathan Arbel @arbel.bsky.social · 26/02/2026

When someone does something wrong, how do we know who did it? When only neighbors were involved, that was hard but solvable. When individuals started cooperating over large distances, or in large groups that shifted members, that became much harder.

100

Yonathan Arbel @arbel.bsky.social · 26/02/2026

In How to Count AIs: Individuation and Liability for AI Agents, with @psalib and philosopher Simon Goldstein, we apply an old legal dilemma to a new problem and offer a time-tested solution. Paper (just posted): papers.ssrn.com/sol3/papers.cf m?abstract_id=6273198

101

Yonathan Arbel @arbel.bsky.social · 26/02/2026

🚨Very soon, billions of AI agents will swarm the economy—copying, splitting, merging at will. Just as soon, someone will get hurt. Is law ready for this moment? We think not quite. But we have a solution.

120

Yonathan Arbel @arbel.bsky.social · 12/08/2025

It's just the first study. Much more work is needed to confirm these results and also evaluate consistency and reliability. It's not clear what aspects of model training contribute the most (RLHF?)

000

Yonathan Arbel @arbel.bsky.social · 12/08/2025

This, alongside other results, is consistent with LLMs having picked up something deep about reasonableness. If true, that would allow tool building for legal and agent applications, with important AI safety implications. but...

100

Yonathan Arbel @arbel.bsky.social · 12/08/2025

Findings: humans judge harshly people who didn't take the socially common level of precautions, even though textbook says that shouldn't matter. Textbook does say they should care about costs, but humans don't care much. LLM repeat this exact pattern in RCTs. (but not all)

100

Yonathan Arbel @arbel.bsky.social · 12/08/2025

To test whether LLMs may have picked up the deeper structure of reasonableness judgment (not just parroting), I offer a new methodology: Silicon RCTs (S-RCTs). Randomize a relevant fact, run session-isolated “participants,” & compare LLM deltas to human deltas.

100

Yonathan Arbel @arbel.bsky.social · 12/08/2025

{A note for CS friends: a large share of law isn’t if-then rules; it’s open-textured standards like “reasonable,” “ordinary meaning,” and “undue burden” But our instruments are thin. Expert intuitions (judges) are criticized as elite, out-of-touch, and crypto-political, juries are noisy&biased}

100

Yonathan Arbel @arbel.bsky.social · 12/08/2025

Two hard problems meet here: (a) courts must infer what ordinary people will reasonably think (e.g., would a teenager read the Pepsi jet as an offer?); (b) we want AI agents to follow open-textured norms (“keep a reasonable distance”)

100

Yonathan Arbel @arbel.bsky.social · 12/08/2025

Reasonableness quietly structures daily life: following distance in traffic, how loud a house party can be, what “up to 50% longer” implies, or whether a Pepsi ad offering a jet is a joke. These judgments are fast, intuitive, and hard to explain. In other words: they are "system-1"

100

Yonathan Arbel @arbel.bsky.social · 12/08/2025

🚨🚨🚨Can judges really know what “reasonable people” think? Can AI help bridge that gap—and can AI agents themselves behave “reasonably” in the wild? A new draft, The Silicon Reasonable Person, asks these questions. tl;dr: early signs point to yes, with careful limits.

110

Reposted by Yonathan Arbel

Kevin Frazier @kevintfrazier.bsky.social · 25/04/2025

While students hit the books, AI professors head to Alabama. Great start to the AI Safety Law Roundtable. It's a real treat to collaborate with @arbel.bsky.social & @petersalib.bsky.social to encourage more profs to evaluate the promise & peril of AI innovation. Stay tuned for more!

011

Yonathan Arbel @arbel.bsky.social · 10/04/2025

I ended up using Walt. I attach the semester plan. I was underwhelmed but it was serviceable

100