The AI Law Review

People writing about computers writing about people.

[The AI Law Review @AILawReview]

Why is generative AI not like a VCR?

Read below to see why the Northern District of California has us asking such a provocative question on a Monday afternoon.

[The AI Law Review @AILawReview]

https://t.co/5gdvuPZjSM

[CDR Magazine @CDR_Magazine]

The Supreme Court of Singapore will impose a string of new AI rules on lawyers and other court users, effective from tomorrow (1 October) – with lawyers facing disciplinary action and financial penalties if they fail to abide by the directions.
https://t.co/wb168bMauJ

[Luiza Jarovsky, PhD @LuizaJarovsky]

🚨 BREAKING: Like in the movies, the judge in one of the AI copyright lawsuits allows authors' representatives to INSPECT OpenAI's confidential TRAINING DATASET in a secure room. Will it become the norm? Here's what will happen:

"Training Data shall be made available for inspection in electronic format at OpenAI’s offices in San Francisco CA, or at a secure location determined by OpenAI within 25 miles of San Francisco, CA; or at another mutually agreed location. Training Data will be made available for inspection between the hours of 8:30 a.m. and 5:00 p.m. on business days, although the parties will be reasonable in accommodating reasonable requests to conduct inspections at other times."

-

"Training Data shall be made available by OpenAI in a secure room on a secured computer without Internet access or network access to other unauthorized computers or devices. The secured computer will contain a README file that will provide a directory of the Training Data and brief descriptions of layout, format, and searching.

-

"No recordable media or recordable devices, including without limitation computers, cellular telephones, cameras, other recording devices, or drives of any kind, shall be permitted into the secure inspection room, except the Producing Party may provide a limited-use note-taking computer at Inspecting Party’s request, solely for note-taking purposes. At the end of each day of inspection, the Inspecting Party shall be able to copy notes from the note-taking computer onto a recordable device, under the supervision of the Producing Party. For the avoidance of doubt, medical devices (e.g., heart monitors, insulin pumps), stopwatches, and timers are permitted so long as such devices are not capable of recording or copying data from the secured computer."

-

"All persons who will review OpenAI’s Training Data on behalf of an Inspecting Party, including the Inspecting Party’s counsel, must qualify under paragraph 7.3 of the Stipulated Protective Order as an individual to whom “HIGHLY CONFIDENTIAL – ATTORNEYS’ EYES ONLY” information may be disclosed, and must sign the Non-Disclosure Agreement attached as Exhibit A to the Stipulated Protective Order."

🔎 Read the full document with the rules applicable to this inspection (link below).

🔥 Interested in AI policy, compliance & regulation? Join 35,400+ people who subscribe to my weekly newsletter (link below).

[TodaysGeneralCounsel @TodaysGC]

Join @TodaysGC and @LegalSuiteNA for an Oct. 1 webinar in which we'll demystify the complex world of Artificial Intelligence (AI) and explore its profound impact on the legal landscape. Don't miss out on the LIVE interactive Q&A session!

Register today: https://t.co/kGtED7odWi

[Jonathan Habshush @jhabshush]

New AI in Law Research: Humans vs. AI in Predicting Case Outcomes

Key Points:

• Dataset: Analyzed 14,582 UK Employment Tribunal cases
• Human Baseline: Two PhD candidates in UK employment law reviewed 1,371 cases
• Prediction Categories: “Claimant wins,” ��Loses,” “Partly wins,” or “Other”
• AI Models Tested: BERT, T5, GPT-3.5, and GPT-4

Fascinating Findings:

• Best AI Model F-score: 56%
• Human Experts F-score: 67%

An F-score measures accuracy by balancing true positives and minimizing false positives, highlighting the practical effectiveness of legal predictions.

Summary:
AI demonstrates strong capabilities in predicting case outcomes, though it still trails behind human experts. Notably, GPT-4 achieved impressive accuracy without any fine-tuning, Retrieval-Augmented Generation (RAG), or Chain of Thought processes. The ablation analysis indicates that GPT-4 correctly predicts about 70% of cases right out-of-the-box, which is remarkable.

Bottom Line:
This study sets a promising baseline rather than an AI ceiling, signaling significant progress for AI in the legal field. With ongoing advancements and the emergence of even more sophisticated models, we are clearly on the right track. This not only enhances AI’s role in legal predictions but also holds great potential for improving access to justice.

https://t.co/qnFHkeGqO5

[New York Law Journal @NYLawJournal]

AI Hiring Poses Discrimination Risk; A Cautionary Tale https://t.co/CwqrA0C7Rh

[Legaltech©news @Legaltech_news]

Tracking Generative AI: How Evolving AI Models Are Impacting Legal https://t.co/0wZDrGEBxp

[The Information @theinformation]

AI Agenda Deal Exclusive: Legal AI startup EvenUp aims to raise its third round of capital in just over a year at a $1 billion price, as VCs bet on vertical AI despite risks.

https://t.co/v0D8vuwuet

From @steph_palazzolo

[Artificial Lawyer @ArtificialLawya]

Litera GenAI Survey: Doc Creation + Proofreading Top Use Cases

https://t.co/jqLR1UXbZJ

[Robert Hosay @robert_hosay]

My colleagues address the new dimensions that #AI introduces to the employment law landscape, and how employers should approach their usage of it. https://t.co/3nHAN8zbnQ

[Ketan Ramakrishnan @ketanrama]

I’m excited to debate SB 1047 and the future of AI regulation with some thoughtful interlocutors this Thursday at @CarnegieEndow - please join (virtually) if you’re free!

[Luiza Jarovsky, PhD @LuizaJarovsky]

🚨 [JOB OPPORTUNITIES] AI Governance IS HIRING! Below are 12 positions in 🇳🇱🇬🇧🇮🇪🇺🇸🇮🇳🇨🇿 posted IN THE LAST FEW DAYS. Bookmark, share & be an early applicant:

[The AI Law Review @AILawReview]

Sharing for anyone interested

[Diego Alcalá Laboy 🇵🇷 @DiegoAlcalaPR]

#Legaltech I teach an Intro to Tech for Lawyers seminar and am looking for 2 guests to join my class to talk about document creation and finalizing documents and e-signatures. If you want to join my class to talk about any of these, reach out!

[Stanford HAI @StanfordHAI]

📢 New policy brief: Could foreign actors use AI to generate persuasive propaganda? Our latest brief presents research that shows the answer is yes – highlighting the need for evidence-based research into the risks posed by LLMs. https://t.co/NZWm8Nnr7t

[Michael Bommarito @mjbommar]

here's the 2nd cir's Hachette opinion from today, which serious implications for fair use analysis in the "AI copyright" cases. buckle up, gang.
https://t.co/BhZh3rwMxw

[Bloomberg Technology @technology]

Microsoft’s investment into Inflection AI avoids in-depth scrutiny from the UK antitrust watchdog after it ruled that the hiring of ex-employees from the startup didn’t cause any regulatory concerns https://t.co/OnTXD61RB1

[Legaltech©news @Legaltech_news]

Deepfakes in Legal Proceedings: A Strategic Framework for Collaborative Solutions https://t.co/isFTfkGn5W

[Litera @Litera_AI]

In today's tech-driven world, the legal sector is rapidly evolving. Litera, @Legaltech_news, and @lawdotcom surveyed law firms on tech adoption, from GenAI to cloud management. Download the report for insights on GenAI, data governance, and cloud trends: https://t.co/4oxZ4oo6n6

[Volokh Conspiracy @VolokhC]

[Eugene Volokh] Fugees Rapper Pras Michel Not Entitled to New Trial Based on Lawyer's Use of AI to Help Craft Closing Argument https://t.co/DV1F9dxl87

[Legaltech©news @Legaltech_news]

Legaltech Rundown: Harvey Partners with Icertis, Flo Recruit Releases Legal Career Platform, and More https://t.co/MT3Qc2m1vA

[Senator Scott Wiener @Scott_Wiener]

On a 29-9 vote, the Senate just gave final sign off to SB 1047, our AI safety bill. The bill now heads to the Governor.