Home

Artificial Intelligence: Using AI to Parse the Logic of a Legal Opinion

Using AI to Parse the Logic of a Legal Opinion

We published an article entitled “Using AI to Parse the Logic of a Legal Opinion.” You can read the article by clicking here

We applied Claude Sonnet 4.6 to the recent Supreme Court tariff case, Learning Resources, Inc. v. Trump, striking down Trump’s tariffs under the statutory authority relied upon by the administration. 

The opinion, with concurrences and dissents, is 170 pages long as a slip opinion. We used AI to parse the logic and assumptions of the multiple opinions. This is an “actual use case” demonstrating the advantages of using AI for legal analysis.

Non-Mutual Offensive Collateral Estoppel Could Not Be Applied By Plaintiffs to Void Arbitration Agreements

Is the Ninth Circuit Ruling Fair?

Nurses sued Aya Healthcare for wage violations. Each nurse had signed an individual arbitration agreement. Four disputes went to separate arbitrations; two arbitrators upheld the agreements, two struck them down as unconscionable. The district court applied non-mutual offensive collateral estoppel — using the two invalidating awards to void 255 other arbitration agreements, without sending those to arbitration. The Ninth Circuit reversed, holding the FAA prohibits this approach. O’Dell, et al. v. Aya Healthcare Services, Inc., No. 25-1528 (9th Cir. 4/1/26) (Tallman, VanDyke, Tung).

Non-mutual offensive collateral estoppel, when it does apply, enables a plaintiff to use collateral estoppel as a sword rather than a shield, applying a helpful outcome from a case in which the plaintiff was not a party to the benefit of the new plaintiff. It is a form of issue preclusion used to avoid relitigating an issue. But unlike classical issue preclusion, it is not mutual because it is used by a party who was not part of the initial litigation, and it is offensive rather than defensive.

At first blush, the outcome seems unjust, because a fundamental principle of justice is that like cases should be decided alike. But here, fairness cuts both ways.

The parties’ consent to arbitrate is a fairness principle, if one can put out of mind the adhesive nature of many employment contracts. Also, two of the four arbitrations accepted the defendant’s position, and one can see a lack of symmetry if issue preclusion were only to be invoked by the plaintiffs here. 

Text-to-Voice Capability Has Been Added To This Blog

We’ve added text-to-voice capability to this website. That means that by clicking on a heading, you should be able to hear the text read out loud. Please let me know if this feature does not work for you. From time to time we will remind readers of this text-to-voice capacity that we added with a plugin.

The Use Of AI Consumer Subscriptions May Compromise Confidentiality

We Published An Article About This In The Daily Journal.

A federal judge in the Southern District of New York has ruled that a client’s legal research using his subscription to an AI platform was not protected by attorney-client privilege, nor by the work-product protection. The client was not an attorney, and the research was not conducted at the direction or behest of the attorney. And the AI program was not an attorney. When a client gives his attorney information that is not protected by attorney-client privilege or work-production protection, it does not convert non-confidential information into confidential information.

We thought that at a time when everyone is turning to AI in order to become their own expert, the confidentiality issue would be of interest to attorneys, mediators, arbitrators, and clients. We are making a copy of our article available here.

Mediation Ethics — An AI Inquiry

An Inquiry Into Mediator Ethics With The Help Of AI

We published an inquiry into mediation ethics with the assistance of AI in 2025 in the last issue of California Litigation, the journal of the Litigation Section of the California Lawyers Association. With the permission of the journal and the CLA, it is republished here and available by inserting the following link in your browser: https://calmediation.org/an-ai-inquiry-into-mediator-ethics-docx-1-1