Does ChatGPT-4 Have Dementia?

Is it just me, or has ChatGPT-4 taken a nosedive when it comes to legal research and writing? There has been a noticeable decline in its ability to locate primary authority on a topic, analyze a fact pattern, and apply law to facts to answer legal questions. Recently, instructions slide through its digital grasp like water through a sieve, and its memory? I would compare it to a goldfish, but I don’t want to insult them. And before you think it’s just me, it’s not just me, the internet agrees!

ChatGPT’s Sad Decline

One of the hottest topics in the OpenAI community, in the aptly named GPT-4 is getting worse and worse every single update thread, is the perceived decline in the quality and performance of the GPT-4 model, especially after the November 2023 update. Many users have reported that the model is deteriorating with each update, producing nonsensical, irrelevant, or incomplete outputs, forgetting the context, and ignoring instructions. Some users have even reverted to previous versions of the model or cancelled their subscriptions. Here are some specific quotations from recent comments about the memory problem:

  • December 2023 – “I don’t know what on Earth is wrong with GPT 4 lately. It feels like I’m talking to early 3.5! It’s incapable of following basic instructions and forgets the format it’s working on after just a few posts.”
  • December 2023 – “It ignores my instructions, in the same message. I can’t be more specific with what I need. I’m needing to repeat how I’d like it to respond every single message because it forgets, and ignores.”
  • December 2023 – “ChatGPT-4 seems to have trouble following instructions and prompts consistently. It often goes off-topic or fails to understand the context of the conversation, making it challenging to get the desired responses.”
  • January 2024 – “…its memory is bad, it tells you search the net, bing search still sucks, why would teams use this product over a ChatGPT Pre Nov 2023.”
  • February 2024 – “It has been AWFUL this year…by the time you get it to do what you want format wise it literally forgets all the important context LOL — I hope they fix this ASAP…”
  • February 2024 – “Chatgpt was awesome last year, but now it’s absolutely dumb, it forgets your conversation after three messages.”

OpenAI has acknowledged the issue and released an updated GPT-4 Turbo preview model, which is supposed to reduce the cases of “laziness” and complete tasks more thoroughly. However, the feedback from users is still mixed, and some are skeptical about the effectiveness of the fix.

An Example of Confusion and Forgetfulness from Yesterday

Here is one of many examples of my experiences which provide an illustrative example of the short-term memory and instruction following issues that other ChatGPT-4 users have reported. Yesterday, I asked it to find some Texas cases about the shopkeeper’s defense to false imprisonment. Initially, ChatGPT-4 retrieved and summarized some relatively decent cases. Well, to be honest, it retrieved 2 relevant cases, with one of the two dating back to 1947… But anyway, the decline in case law research ability is a subject for another blog post.

Anyway, in an attempt to get ChatGPT-4 to find the cases on the internet so it could properly summarize them, I provided some instructions and specified the format I wanted for my answers. Click here for the transcript (only available to ChatGPT-4 subscribers).

Confusion ran amok! ChatGPT-4 apparently understood the instructions (which was a positive sign) and presented three cases in the correct format. However, they weren’t the three cases ChatGPT had listed; instead, they were entirely irrelevant to the topic—just random criminal cases.

It remembered… and then forgot. When reminded that I wanted it to work with the first case listed and provided the citation, it apologized for the confusion. It then proceeded to give the correct citation, URL, and a detailed summary, but unfortunately in the wrong format!

Eventually, in a subsequent chat, I successfully got it to take a case it found, locate the text of the case on the internet, and then provide the information in a specified format. However, it could only do it once before completely forgetting about the specified format. I had to keep cutting and pasting the instructions for each subsequent case.

Sigh… I definitely echo the sentiments of expressed on the GPT-4 is getting worse and worse every single update thread.

ChatGPT Is Growing a Long Term Memory

Well, the news is not all bad! While we are on the topic of memory, OpenAI has introduced a new feature for ChatGPT – the ability to remember stuff over time. ChatGPT’s memory feature is being rolled out to a small portion of free and Plus users, with broader availability planned soon. According to OpenAI, this enhancement allows ChatGPT to remember information from past interactions, resulting in more personalized and coherent conversations. During conversations, ChatGPT automatically picks up on details it deems relevant to remember. Users can also explicitly instruct ChatGPT to remember specific information, such as meeting note preferences or personal details. Over time, ChatGPT’s memory improves as users engage with it more frequently. This memory feature could be useful for users who want consistent responses, such as replying to emails in a specific format.

The memory feature can be turned off entirely if desired, giving users control over their experience. Deleting a chat doesn’t erase ChatGPT’s memories; users must delete specific memories individually…which seems a bit strange – see below. For conversations without memory, users can use temporary chat, which won’t appear in history, won’t use memory, and won’t train the AI model.

The Future?

As we await improvements to our once-loved ChatGPT-4, our options remain limited, pushing us to consider alternative avenues. Sadly, I’ve encountered recent similar shortcomings with the once-useful for legal research and writing Claude 2. In my pursuit of alternatives, platforms like Gemini, Perplexity, and Hugging Face have proven less than ideal for research and writing tasks. However, amidst these challenges, Microsoft Copilot has shown promise. While not without its flaws, it recently demonstrated adequate performance in legal research and even took a passable stab at a draft of a memo. Given OpenAI’s recent advancements in the form of Sora, the near-magical text-to-video generator that is causing such hysteria in Hollywood, there’s reason to hope that they can pull ChatGPT back from the brink.

ABA TECHSHOW 2024 Review

Since so many of the AI Law Librarians team were able to attend this year, we thought we would combine some of our thoughts (missed you Sarah!) about this yearly legal technology conference.


Startup Alley

We arrived in Chicago on a chilly Wednesday morning, amid an Uber & Lyft strike, with plenty of time to take the train from the airport to our hotel. After an obligatory trip to Giordanno’s our students were ready to head over to the Start-up Pitch Competition. I sat with co-blogger Rebecca Fordon during the competition and we traded opinions on the merits of the start-up pitches. We both come from the academic realm and were interested in seeing the types of products that move the needle for attorneys working at firms.

I was familiar with many of the products because I spend a decent portion of my time demo’ing legal tech as part of my current role. It was stiff competition and there were many outstanding options to choose from. Once all of the pitches were done, the audience voted, and then Bob Ambrogi announced the winners. To my great surprise and pleasure, AltFee won! For the uninitiated, AltFee is “a product that helps law firms replace the billable hour with fixed-fee pricing.” This was very interesting to me because I have long thought that LLMs could mean the death knell of the billable hour in certain legal sectors. This was, at least, confirmation that the attorneys attending the TECHSHOW have this on their radar and are thinking through how they are going to solve this problem.

techshow sessions

This year’s schedule of sessions was noticeably heavy on AI-related topics. This was great for me because I’m super interested in this technology and how it is being implemented in the day-to-day life of practitioners. I saw sessions on everything from case management software, to discovery, to marketing, kinda everything.

An especially inspiring couple of sessions for me featured Judge Scott Schlegel on the Fifth Circuit Court of Appeal in Louisiana. Judge Schlegel is the first judge that I’ve seen make fantastic use of AI in United States Courts for access to justice. I am passionate about this topic and have been fishing for grants to try to implement a handful of projects that I have so it was phenomenal to see that there are judges out there who are willing to be truly innovative. Any initiative for access to justice in the courts would require the buy-in of many stakeholders so having someone like Judge Schlegel to point to as a proof of concept could be crucial in getting my projects off the ground. After hearing his presentations I wished that every court in the US had a version of him to advocate for these changes. Importantly, none of his projects require tons of funding or software development. They are small, incremental improvements that could greatly help regular people navigate the court system – while, in many cases, improving the daily lives of the court staff and judges who have to juggle huge caseloads. Please feel free to email grants opportunities in this vein if you see them:

side quest: northwestern law ai symposium

In the weeks leading up to the TECHSHOW I received an invite from Prof. Daniel Linna to attend Northwestern University’s AI and Law: Navigating the Legal Landscape of Artificial Intelligence Symposium. I took a frigid hike down to the school in the morning to attend a few sessions before returning to the TECHSHOW in the afternoon. It was a fantastic event with a great mix of attorneys, law professors, and computer science developers.

I was able to see Professor Harry Surden‘s introductory session on how LLM’s work in legal applications. While this information was not “new” to me per se (since I frequently give a similar presentation), he presented this complicated topic in an engaging, clear, and nuanced way. He’s obviously a veteran professor and expert in this area and so his presentation is much better than mine. He gave me tons of ideas on how to improve my own presentations to summarize and analogize these computer science topics to legal professionals, for which I was very grateful.

The second session was a panel that included Sabine BrunswickerJJ Prescott, and Harry Surden. All were engaged in fascinating projects using AI in the law and I encourage you to take a look through their publications to get a better sense of what the pioneers in our field are doing to make use of these technologies in their research.

Our Students

Each year our school funds a cohort of students to attend the TECHSHOW and this year was no different. This is my first year going with them and I wasn’t sure how much value they would get out of it since they don’t have a ton of experience working in firms using these tools. Was this just a free trip to Chicago or was this pedagogically useful to them?

I will cut to the chase and say that they found this tremendously useful and loved every session that they attended. Law school can (sometimes) get a little disconnected from the day-to-day practice of law and this is a great way to bridge that gap and give the students a sense of what tools attorneys use daily to do their jobs. You’d think that all of the sexy AI-related stuff would be attractive to students but the best feedback came from sessions on basic office applications like MS Outlook and MS Word. Students are definitely hungry for this type of content if you are trying to think through workshops related to legal technology.

In addition to the sessions, the students greatly appreciated the networking opportunities. The TECHSHOW is not overly stuffy and formal and I think they really liked the fact that they could, for example, find an attorney at a big firm working in M&A and pick their brain at an afterparty to get the unfiltered truth about a specific line of work. All of the students said they would go again and I’m going to try to find ways to get even more students to attend next year. If your school ends up bringing students in the future, please reach out to me and we can have our students get together at the event.


Jenny live-tweeted the ABA TECHSHOW’s 60 Apps in 60 Minutes and provided links. You can follow her on this exciting journey starting with this tweet:


One of the most impactful sessions for me was titled “Revitalize Your Law Firm’s Knowledge Management with AI,” with Ben Schorr (Microsoft) and Catherine Sanders Reach (North Carolina Bar Association).  To drive home why KM matters so much, they shared the statistic that knowledge workers spend a staggering 2.5 hours a day just searching for what they need. That resonated with me, as I can recall spending hours as a junior associate looking for precedent documents within my document management system. Even as a librarian, I often spend time searching for previous work that either I or a colleague has done.

To me, knowledge management is one of the most exciting potential areas to apply AI, because it’s such a difficult problem that firms have been struggling with for decades. The speaker mentioned hurdles like data silos (e.g., particular practice areas sharing only among themselves), a culture of hoarding information, and the challenges of capturing and organizing vast amounts of data, such as emails and scanned documents with poor OCR. 

The speakers highlighted several AI tools that are attempting to address these issues through improved search going beyond keywords, automating document analysis to aid in categorizing documents, and suggesting related documents. They mentioned Microsoft Copilot, along with process tools like Process Street, Trainual, and Notion. Specific tools like Josef allow users to ask questions of HR documents and policies, rather than hunting for the appropriate documents.

Artificial Intelligence and the Future of Law Libraries Roundtable Events

South Central Roundtable

OU Law volunteered to host the South Central “Artificial Intelligence and the Future of Law Libraries” roundtable and so I was fortunate enough to be allowed to attend. This is the third iteration of a national conversation on what the new AI technologies could mean for the future of law libraries and (more broadly) law librarianship. I thought I would fill you in on my experience and explain a little about the purpose and methodology of the event. The event follows Chatham House Rules so I cannot give you specifics about what anybody said but I can give you an idea of the theme and process that we worked through.

Law Library Director Kenton Brice of OU Law elected to partner with Associate Dean for Library and Technology Greg Ivy and SMU to host the event in Dallas, TX because it was more accessible for many of the people that we wanted to attend. I’d never been to SMU and it’s a beautiful campus in an adorable part of Dallas – here’s a rad stinger I made in Premiere Pro:

Not cleared with SMU’s marketing department

TL;DR: If you get invited, I would highly recommend that you go. I found it enormously beneficial.

History and Impetus

The event is the brainchild of Head of Research, Data & Instruction, Director of Law Library Fellows Program Technology & Empirical Librarian, Cas Laskowsi at the University of Arizona (hereinafter “Cas”). They hosted the inaugural session through U of A’s Washington, DC campus. You may have seen the Dewey B. Strategic article about it since Jean O’Grady was in attendance. The brilliant George H. Pike at Northwestern University hosted the second in the series in Chicago. I know people who have attended each of these sessions and the feedback has been resoundingly positive.

The goal of this collaborative initiative is to provide guidance to law libraries across the country as we work to strategically incorporate artificial intelligence into our operations and plan for the future of our profession. 

Cas, from the U of A Website


The event takes the entire day and it’s emotionally exhausting, in the best way possible. We were broken into tables of 6 participants. The participants were hand-selected based on their background and experience so that each table had a range of different viewpoints and perspectives.

Then the hosts (in our case, Kenton Brice and Cas Laskowski) walked us through a series of “virtuous cycle, vicious cycle” exercises. They, thankfully, started with the vicious cycle so that you could end each session on a virtuous cycle, positive note. At the end, each table chose a speaker and then we summarized the opinions discussed so that the entire room could benefit from the conversations. Apparently, this is an exercise done at places like the United Nations to triage and prepare for future events. This process went on through 3 full cycles and then we had about an hour of open discussion at the end. We got there at 8am and had breakfast and lunch on-site (both great – thank you Greg Ivy and SMU catering) because it took the entire day.

We had a great mix of academic, government, and private sector presented at the event and the diversity of stakeholders and experiences made for robust and thought-provoking conversation. Many times I would hear perspectives that had never occurred to me and would have my assumptions challenged to refine my own ideas about what the future might look like. Additionally, the presence of people with extensive expertise in specific domains, such as antitrust, copyright, the intricacies of AMLaw100 firms, and the particular hurdles faced in government roles, enriched the discussions with a depth and nuance that is rare to find. Any one of these areas can require years of experience so having a wide range of experts to answer questions allowed you to really “get into the weeds” and think things through thoroughly.

My Experience

I tend to be (perhaps overly) optimistic about the future of these technologies and so it was nice to have my optimism tempered and refined by people who have serious concerns about what the future of law libraries might look like. While the topics presented were necessarily contentious, everybody was respectful and kind in their feedback. We had plenty of time for everybody to speak (so you didn’t feel like you were struggling to get a word in).

You’d think that 8 hours of talking about these topics would be enough but we nearly ran over on every exercise. People have a lot of deep thoughts, ideas, and concerns about the state and future of our industry. Honestly, I would have been happy to have this workshop go on for several days and cover even more topics if that was possible. I learned so much and gained so much value from the people at my table that it was an incredibly efficient way to get input and share ideas.

Unlike other conferences and events that I’ve attended this one felt revolutionary – as in, we truly need to change the status quo in a big way and start getting to work on new ways to tackle these issues. “Disruptive” has become an absolute buzzword inside of Silicon Valley and academia but now we have something truly disruptive and we need to do something about it. Bringing all these intelligent people together in one room fosters an environment where disparate, fragmented ideas can crystalize into actionable plans, enabling us to support each other through these changes.

The results from all of these roundtables are going to be published in a global White Paper once the series has concluded. Each roundtable has different regions and people involved and I can’t wait to see the final product and hear what other roundtables had to say about these important issues. More importantly, I can’t wait to be involved in the future projects and initiatives that this important workshop series creates.

I echo Jean O’Grady: If you get the call, go.

Birth of the Summarizer Pro GPT: Please Work for Me, GPT

Last week, my plan was to publish a blog post about creating a GPT goofily self-named Summarizer Pro to summarize articles and organize citation information in a specific format for inclusion in a LibGuide. However, upon revisiting the task this week, I find myself first compelled to discuss the recent and thrilling advancements surrounding GPTs – the ability to incorporate GPTs into a ChatGPT conversation.

What is a GPT?

But, first of all, what is a GPT? The OpenAI website explains that GPTs are specialized versions of ChatGPT designed for customized applications. These unique GPTs enable anyone to modify ChatGPT for enhanced utility in everyday activities, specific tasks, professional environments, or personal use, with the added ability to share these personalized versions with others.

To create or use a GPT, you need access to ChatGPT’s advanced features, which require a paid subscription. Building your own customized GPT does not require programming skills. The process involves starting a chat, giving instructions and additional information, choosing capabilities like web searching, image generation, or data analysis, and iteratively testing and improving the GPT. Below are some popular examples that ChatGPT users have created and shared in the ChatGPT store:

GPT Mentions

This was already exciting, but last week they introduced a feature that takes it to the next level – users can now invoke a specialized GPT within a ChatGPT conversation. This is being referred to as “GPT mentions” online. By typing the “@” symbol, you can choose from GPTs you’ve used previously for specific tasks. Unfortunately, this feature hasn’t rolled out to me yet, so I haven’t had the chance to experiment with it, but it seems incredibly useful. You can chat with ChatGPT as normal while also leveraging customized GPTs tailored to particular needs. For example, with the popular bots listed above, you could ask ChatGPT to summon Consensus to compile articles on a topic. Then call on Write For Me to draft a blog post based on those articles. Finally, invoke Image Generator to create a visual for the post. This takes the versatility of ChatGPT to the next level by integrating specialized GPTs on the fly.

Back to My GPT Summarizer Pro

Returning to my original subject, which is employing a GPT to summarize articles for my LibGuide titled ChatGPT and Bing Chat Generative AI Legal Research Guide. This guide features links to articles along with summaries on various topics related to generative AI and legal practice. Traditionally, I have used ChatGPT (or occasionally Bing or Claude 2, depending on how I feel) to summarize these articles for me. It usually performs admirably well on the summary part, but I’m left to manually insert the title, publication, author, date, and URL according to a specific layout. I’ve previously asked normal old ChatGPT to organize the information in this format, but the results have been inconsistent. So, I decided to create my own GPT tailored for this task, despite having encountered mixed outcomes with my previous GPT efforts.

Creating GPTs is generally a simple process, though it often involves a bit of fine-tuning to get everything working just right. The process kicks off with a set of questions… I outlined my goals for the GPT – I needed the answers in a specific format, including the title, URL, publication name, author’s name, date, and a 150-word summary, all separated by commas. Typically, crafting a GPT involves some back-and-forth with the system. This was exactly my experience. However, even after this iterative process, the GPT wasn’t performing exactly as I had hoped. So, I decided to take matters into my own hands and tweak the instructions myself. That made all the difference, and suddenly, it began (usually) producing the information in the exact format I was looking for.

Summarizer Pro in Action!

Here is an example of Summarizer Pro in action! I pasted a link to an article into the text box and it produced the information in the desired format. However, reflecting the dynamic nature of ChatGPT responses, the summaries generated this time were shorter compared to last week. Attempts to coax it into generating a longer or more detailed summary were futile… Oh well, perhaps they’ll be longer if I try again tomorrow or next week.

Although it might not be the most fancy or thrilling use of a GPT, it’s undeniably practical and saves me time on a task I periodically undertake at work. Or course, there’s no shortage of less productive, albeit entertaining, GPT applications, like my Ask Sarah About Legal Information project. For this, I transformed around 30 of my blog posts into a GPT that responds to questions in the approximate manner of Sarah.

Is Better Case Law Data Fueling a Legal Research Boom?

Recently, I’ve noticed a surge of new and innovative legal research tools. I wondered what could be fueling this increase, and set off to find out more. 

The Moat

An image generated by DALL-E, depicting a castle made of case law reporters, with sad business children trying to construct their own versions out of pieces of paper. They just look like sand castles.

Historically, acquiring case law data has been a significant challenge, acting as a barrier to newcomers in the legal research market. Established players are often protective of their data. For instance, in an antitrust counterclaim, ROSS Intelligence accused Thomson Reuters of withholding their public law collection, claiming they had to instead resort to purchasing cases piecemeal from sources like Casemaker and Fastcase.  Other companies have taken more extreme measures. For example, Ravel Law partnered with the Harvard Law Library to scan every single opinion in their print reporter collections. There’s also speculation that major vendors might even license some of their materials directly to platforms like Google Scholar, albeit with stringent conditions.

The New Entrants

Despite the historic challenges, several new products have recently emerged offering advanced legal research capabilities:

  • (founded 2023) – This platform leverages generative AI to read and summarize judicial opinions, streamlining the search process. Currently hosting around 1.6 million summarized opinions, it’s available for free.
  • Midpage (2022) – Emphasizing the integration of legal research into the writing process, users can employ generative AI to draft documents from selected source (see Nicola Shaver’s short writeup on Midpage here). Midpage is currently free at
  • CoPilot (by LawDroid, founded 2016) – Initially known for creating chatbots, LawDroid introduced CoPilot, a GPT-powered AI legal assistant, in 2023. It offers various tasks, including research, translating, and summarizing. CoPilot is available in beta as a web app and a Chrome extension, and is free for faculty and students.
  • (2023) – Another generative AI legal assistant, allows users to conduct legal research, draft documents, and more. Limited free access is available without signup at, although case law research will require you to sign up for a free account.
  • Alexi (2017) Originally focused on Canadian law, Alexi provides legal research memos. They’ve recently unveiled their instant memos, powered by generative AI. Alexi is available at and provides a free pilot.

Caselaw Access Project and Free Law Project

With the Caselaw Access Project, launched in 2015, Ravel Law and Harvard Law Library changed the game. Through their scanning project, Harvard received rights to the case law data, and Ravel gained an exclusive commercial license for 8 years. (When Lexis acquired Ravel a few years later, they committed to completing the project.) Although the official launch date of free access is February 2024, we are already seeing a free API at Ravel Law (as reported by Sarah Glassmeyer).

Caselaw Access Project data is only current through 2020 (scanning was completed in 2018, and has been supplemented by Fastcase donations through 2020) and does not include digital-first opinions. However, this gap is mostly filled through CourtListener, which contains a quite complete set of state and federal appellate opinions for recent years, painstakingly built through their network of web scrapers and direct publishing agreements. CourtListener offers an API (along with other options for bulk data use).

And indeed, Caselaw Access Project and Free Law Project just recently announced a dataset called Collaborative Open Legal Data (COLD) – Cases. COLD Cases is a dataset of 8.3 million United States legal decisions with text and metadata, suitable for use in machine learning and natural language processing projects.

Most of the legal research products I mentioned above do not disclose their precise source of their case law data. However, both and Midpage point to CourtListener as a partner. My theory/opinion is that many of the others may be using this data as well, and that these new, more reliable and more complete sources of data are responsible for fueling some amazing innovation in the legal research sphere.

What Holes Remain?

Reviewing the coverage of CourtListener and Caselaw Access Project it appears to me that they have, when combined:

  • 100% of all published U.S. case law from 2018 and earlier (state and federal)
  • 100% of all U.S. Supreme Court, U.S. Circuit Court of Appeals, and state appellate court cases

There are, nevertheless, still a few holes that remain in the coverage:

  • Newer Reporter Citations. Newer appellate court decisions may not have reporter citations within CourtListener. These may be supplemented as Fastcase donates cases to Caselaw Access Project.
  • Newer Federal District Court Opinions. Although CourtListener collects federal decisions marked as “opinions” within PACER, these decisions are not yet available in their opinion search. Therefore, very few federal district court cases are available for the past 3-4 years. This functionality will likely be added, but even when it is, district courts are inconsistent about marking decisions as “opinions” and so not all federal district court opinions will make their way to CourtListener’s opinions database. To me, this brings into sharp relief the failure of federal courts to comply with the 2002 E-Government Act, which requires federal courts to provide online access to all written opinions.
  • State Trial Court Decisions. Some other legal research providers include state court trial-level decisions. These are generally not published on freely available websites (so CourtListener cannot scrape them) and are also typically not published in print reporters (so Caselaw Access Project could not scan them).
  • Tribal Law. Even the major vendors have patchy access to tribal law, and CourtListener has holes here as well.

The Elephant in the Room

Of course, another major factor in the increase in legal research tools may be simple economics. In August, Thomson Reuters acquired the legal research provider Casetext for the eye-watering sum of $650 million.  And Casetext itself is a newer legal research provider, founded only in 2013. In interviews, Thomson Reuters cited Casetext’s access to domain-specific legal authority, as well as its early access to GPT-4, as key to its success. 

What’s Next?

Both Courtlistener and Caselaw Acess Project have big plans for continuing to increase access to case law. CAP will launch free API access in February 2024, coordinating with LexisNexis, Fastcase, and the Free Law Project on the launch. CourtListener is planning a scanning project to fix remaining gaps in their coverage (CourtListener’s Mike Lissner tells me they are interested in speaking to law librarians about this – please reach out). And I’m sure we can expect to see additional legal research tools, and potentially entire LLMs (hopefully open source!), trained on this legal data.

Know of anything else I didn’t discuss? Let me know in the comments, or find me on social media or email.

Beware the Legal Bot: Spooky Stories of AI in the Courtroom

The “ChatGPT Attorney” case has drawn much attention, but it’s not the only example of lawyers facing problems with AI use. This blog will compile other instances where attorneys have gotten into trouble for incorporating AI into their practice. Updates will be made as new cases or suggestions arise, providing a centralized resource for both legal educators and practicing attorneys (or it can be used to update a Libguide 😉). I’ll will also add this to one of our menus or headings for easy access.

Attorney Discipline 

Park v. Kim, No. 22-2057, 2024 WL 332478 (2d Cir. Jan. 30, 2024)

“Attorney Jae S. Lee. Lee’s reply brief in this case includes a citation to a non-existent case, which she admits she generated using the artificial intelligence tool ChatGPT. Because citation in a brief to a non-existent case suggests conduct that falls below the basic obligations of counsel, we refer Attorney Lee to the Court’s Grievance Panel, and further direct Attorney Lee to furnish a copy of this decision to her client, Plaintiff-Appellant Park.”

Mata v. Avianca, Inc. (1:22-cv-01461) District Court, S.D. New York 

I will not belabor the ChatGPT attorney (since it has been covered by real journalists like the NYT) – only provide links to the underlying dockets in case you need them since I get asked for them fairly often:

(Fireworks start at the May 4, 2023 OSC)

Zachariah Crabhhill, Colorado Springs 

In a less publicized case from Colorado, an attorney, Zachariah Crabhill, relied on ChatGPT to draft a legal motion, only to find out later that the cited cases were fictitious. Unfortunately, the court filings are not accessible through El Paso County’s records or Bloomberg Law. If any Colorado law librarians can obtain these documents, please contact me, and I’ll update this post accordingly.

News articles: 

Zachariah was subsequently sanctioned and suspended:

Ex Parte Allen Michael Lee, No. 10-22-00281-CR, 2023 WL 4624777 (Tex.
Crim. App. July 19, 2023)

An Opinion of Chief Justice Tom Grey explains that Allen Michael Lee faces charges related to child sexual assault, with bail set at $400,000, which he hasn’t been able to post. Lee sought a bail reduction through a pre-trial habeas corpus application, but the court denied this, leading Lee to argue that the denial was an abuse of discretion due to excessive initial bail. However, his appeal was critiqued for inadequate citation, as the cases he referenced either didn’t exist or were unrelated to his arguments


David Wagner, This Prolific LA Eviction Law Firm Was Caught Faking Cases In Court. Did They Misuse AI?, LAist (Oct 12, 2023)
Submitted by my co-author Rebecca Fordon

Cuddy Law Firm in New York has been submitting exhibits of transcripts of interactions with ChatGPT to their motions for attorneys fees (essentially a back and forth to zero in on what is a reasonable rate) in several cases in S.D. NY.”
[This is an ongoing action and we’re waiting to see if it is allowed]
from reader Jason as a comment (very much appreciated, Jason)

A Spooky Glimpse into the Future

In 2019, Canadian Judge Whitten reduced an attorney’s requested fees on the grounds that the attorney had not utilized AI technology:

The decision concerned a request for attorneys’ fees and expenses by defendant, Port Dalhousie Vitalization Corporation (PDVC). The court granted summary judgment in PDVC’s favor against a woman who sued PDVC after she slipped and fell at an Ontario bar for which PDVC was the landlord. The bar, My Cottage BBQ and Brew, defaulted in the case. In his ruling, Justice Whitten mentioned that the use of AI in legal research would have reduced the amount of time one of the attorneys for the defendant would have spent preparing his client’s case.

In domains where AI can significantly expedite workflows, it could indeed become standard practice for judges to scrutinize fee requests more rigorously. Attorneys might be expected to leverage the latest technological tools to carry out tasks more efficiently, thereby justifying their fees. In this scenario, sticking to traditional, manual methods could be perceived as inefficient, and therefore, not cost-effective, leading to fee reductions. This has led many people to wonder if AI will expedite the decline of the billable hour (for more on that please see this fantastic discussion on 3 Geeks and a Law Blog, AI-Pocalypse: The Shocking Impact on Law Firm Profitability).

We hope that you have a Happy Halloween!

Crystal Ballalytics: Judicial Behavioral Forecasting Modeling

The trifecta of big data, advanced analytics, and recent AI innovations is ushering in a new era of judicial analytic mind-reading, enabling software to more accurately predict judges’ court rulings. Last year, in what seems like an interesting leap forward in judicial analytics, Pre/Dicta unveiled its AI-powered litigation prediction software, introducing a novel, perhaps radical, approach to tapping into the judicial mind. According to CEO Dan Rabinowitz, Pre/Dicta is the only litigation analytics platform that makes verifiable predictions about the outcome of lawsuits. He claims that using data science and only a docket number, Pre/Dicta’s software correctly forecasts how judges will decide on motions to dismiss 86% of the time, without factoring in the specific facts of the case. The system covers civil litigation cases at both the state and federal level, but does not attempt to forecast results of jury trials.

Rather than solely depending on a judge’s past rulings and jurisprudence, as is common with other judicial analytics products, Pre/Dicta uses a methodology similar to that used in targeted advertising. This approach forecasts future behavior by examining both past actions, such as purchasing habits, and individual biographical characteristics. Pre/Dicta works by combining historical ruling data with biographical and demographic details to forecast a judge’s decision in a given case. Using around 120 data points, it spots patterns and potential biases in a judge’s past rulings. The system evaluates specifics of past rulings, considering elements such as the nature of the case (e.g., securities fraud, employment discrimination), the attorneys and firms involved (e.g., solo practitioner representing an individual, regional firm representing a corporation, AmLaw 100 firm backing an individual), and the nature of the disputing parties (e.g., individual vs. corporation, small company vs. large corporation). This case-specific information is then combined with the judge’s personal data, like net worth, political affiliations, professional history, and law school alma mater, to generate a prediction.

Prediction in the Legal Landscape

86% accuracy is impressive! Hopefully, Pre/Dicta will spark a judicial prediction analytics arms race. According to Daniel L. Chen in his article, “Judicial Analytics and the Great Transformation of American Law,” predictive judicial analytics “holds the promise of increasing the efficiency and fairness of law.” Targeted advertising seems to work pretty well, so hopefully Pre/Dicta’s advancements in this area is a positive step toward making the judicial process more transparent.

If only we knew what would happen in the future, we would know what to do now! For as long as there have been courts and judges, folks have tried to predict whether a judge would rule in their favor. Attorneys have always engaged in mental “judicial analytics” by gathering and pondering information on a judge’s past rulings and reputation to glean some insights into how they might decide a case. Humans are prediction machines, given our innate tendency to draw on experiences and knowledge to anticipate events—an evolutionarily useful skill that allowed us to sometimes escape being saber-toothed tiger lunch or the victim of grumpy neighboring tribal predations.

From my brief stint practicing family law in the 1990s, I discovered that family law clients are hopeful individuals. Despite clear child support guidelines and a prevailing judicial preference for shared custody, people often believed that if a judge merely heard the specifics of their “special snowflake” scenario involvinga cheating spouse or a deadbeat dad, the judge would surely deviate from the rules and customary practices to grant them a deserved favorable ruling. They struggled to accept that judges could be indifferent to their parade of marital/parental horribles. And even if judges were initially inclined to empathize, after many years of sifting through outright lies and half-truths, they had seemingly given up on given up on deciphering reality anyway. It was always challenging to persuade clients of the judicial propensity to metaphorically split the baby down the middle, whether financially or custodially.

Attorneys have needed to hone their abilities to predict outcomes so they could counsel their clients on different courses of action. While making no promises, they share predictions regarding claim values, the odds of surviving summary judgment, potential jail sentences, the likelihood of obtaining sole custody of children, and so on. Attorneys can only do so much, though. Hopefully, as predictive judicial analytics tools improve and become widely available, they have the potential to promote fairness, cut down on litigation costs, and create a more transparent and predictable judicial system.

Judicial Behavioral Forecasting Modeling

Certainly, judges do provide clients with information that assists in anticipating how a ruling might unfold. I have observed numerous judges delivering impactful speeches during temporary hearings, highlighting the importance of adhering to child support guidelines and the principle of shared custody. When clients receive information regarding a likely outcome, their acceptance of reality accelerates significantly. It would indeed be beneficial, and save a lot of time, money, and anguish, if a client could engage in a comprehensive discussion with a judge, probing various questions about how different pieces of information might influence their ruling. However, this isn’t the modus operandi of judges, as they cannot communicate with one party in a suit independently, nor do they pre-announce their rulings prior to a hearing or trial. Now, however, companies like Pre/Dicta are leveraging the human trait of predictability inherent in judges. Like everyone, judges have their own set of ideas, habits, preferences, prejudices, and temperaments shaped by a mix of genetics and experiences, all of which contribute to a certain level of predictability in their rulings.

Hopefully, soon we will be able to pick the mind of a judge without the necessity of actually speaking with her. With the advancing tide of artificial intelligence and the ongoing proliferation and refinement of judicial analytics products, it seems plausible that the future might produce a family law judge behavioral forecasting model for specific judges. These models could help attorneys and their clients identify potential biases of judges. They could see how a judge might respond to a person based on certain characteristics like sex, race, age, income, profession, or criminal history, especially when compared to another party with a different background. Also, if these models included information about factors that affected past rulings, they could be used to anticipate how certain situations might be viewed by the court. For example, a parent hoping to keep their soon-to-be ex-spouse away from the kids might want to know if the judge objects to stuff like dating app addiction, not taking the child to piano lessons, or multiple DUIs arrests. Armed with information, they can choose the best way to handle their case, including deciding if going to trial is a good idea.

Behavioral forecasting models are of course not new to law and legal practice. They are tools used to predict the likely behaviors of individuals or groups across various domains, aiding in better decision-making. In the legal sector, in addition to predicting the outcome of Supreme Court cases, they aid in litigation strategy, legal analytics, resource allocation, criminal behavior prediction, policy impact analysis, legal document analysis, dispute resolution, and regulatory compliance, by leveraging historical data and legal precedents to inform decision-making and strategy development. They are utilized in other fields too like marketing, finance, HR, healthcare, public policy, urban planning, criminal justice, technology, environmental science, and education to forecast behavioral patterns, helping to optimize strategies and allocate resources more efficiently.

Such an innovation would undeniably be a game changer. Clients in divorce and custody disputes might believe solid advice regarding likely outcomes, rather than cling to the hope that their unique case details will influence the judge. Accurate predictions would likely deter individuals from wasting money, and likely be a boon for judges struggling with a backlog of cases. Having these predictive tools on their websites would no doubt promote case settlements and therefore ease some of the strain on both judges and the judicial system.


As always, naysayers abound. Some argue that judicial analytics could undermine the legitimacy of an impartial judiciary. In fact, in France, judges are so wary of transparency that judicial analytics products are prohibited. Well, at least in the U.S., that boat has sailed, far and fast. Particularly in light of Supreme Court rulings in recent years, many people have realized that judges often base their rulings on ideological leanings and personal preferences. Robots would only further confirm what we already suspect – that judges are just like the rest of us with habits, biases, and opinions. It might be too late to rehabilitate the judiciary, but perhaps the transparency of data-driven prediction could bolster public confidence more than frequent affirmations of judicial objectivity.

Then, there are arguments regarding fairness due to cost. For now, the high cost of Pre/Dicta raises potential fairness issues, as only larger firms and wealthier clients can harness its predictive power. True, they always have an advantage. However, as the technology becomes more common, costs should decrease, making it more and more accessible.


The improvement of AI-driven judicial analytics, exemplified by Pre/Dicta, could mark a revolutionary shift in the legal realm, perhaps promising a new level of predictability and transparency in court outcomes. While concerns about fairness, accessibility, and the perception of judicial impartiality persist, the potential benefits—reduced litigation costs, enhanced transparency, and more informed decision-making—may herald a future where data-driven insights guide legal strategy and expectations. As technology continues to evolve and become more accessible, the future looks promising for both practitioners and those seeking justice.

Keeping Up With Generative AI in the Law

The pace of generative AI development (and hype) over the past year has been intense, and difficult even for us experienced librarians, masters of information that we are, to follow. Not only is there a constant stream of new products, but also new academic papers, blog posts, newsletters, and more, from people evaluating, experimenting with, and critiquing those products. With that in mind, I’m sharing my favorites, and I’ll also pepper in a few recommendations from my co-bloggers.


Before Twitter began its slow decline, it was one of my primary sources for professional connection, and there are many there who are exploring generative AI. I especially enjoy following people outside of the legal world. Many of my favorites are still there, like Ethan Mollick, Anna Mills, and Lance Eaton (all in higher education) as well as critical AI theorists like Timnit Gibru and Emily Bender.


Despite the good bits that remain on Twitter, many interesting legal tech discussions seem to have moved to LinkedIn (or perhaps I’ve only recently found them there). Some of my favorites to follow on LinkedIn (in no particular order beyond how I’m running across them as I scroll) are: Nicole Black, Sam Harden, Alex Smith, Cat Moon, Damien Riehl, Dennis Kennedy, Uwais Iqbal, Ivy Grey, Robert Ambrogi, Cat Casey, Nicola Shaver, Adam Ziegler, and Michael Bommarito. Both Bob Ambrogi and Nicola Shaver recently had posts gathering legal tech luminaries to follow, so I would recommend checking out those posts and the comments to find more interesting folks. And if anyone else has figured out the LinkedIn etiquette for connecting vs. following someone you only know via other social media, please let me know.


Most of us have many (many, many) newsletters filling our inbox each day. Here are some favorites.


  • AI in Education – a Google group
  • Lawyer Ex Machina – from law librarian Eli Edwards, on legal technology, law practice and selected issues around big data, artificial intelligence, blockchain, social media and more affecting both the substance and the business of law (weekly)
  • The Neuron – AI news, tools, and how-to
  • The Brainyacts – from Josh Kubicki, insight & tips on generative AI use in legal services (daily)


  • One Useful Thing – from Ethan Mollick, mostly on AI in higher ed (weekly)
  • Do Something – from Sam Harden, on legal tech, often from a small firm and access to justice angle
  • Legal Tech Trends – legal tech links, podcast, articles, products, along with original pieces (every two weeks or so)
  • KnowItAALL – this daily newsletters is a benefit for members of AALL (American Association of Law Libraries), but it is also available to non-members for a fee; great coverage of legal AI, I read it every day
  • AI Law Librarians – is it gauche to recommend our own blog? You can subscribe as a newsletter if you like!



There are loads of podcasts on AI, but here are a few we follow:

Blogs & Websites

We’re bloggers, we like blogs. Traditional media can be ok, too, although mind the paywall.


Sean also mentioned that much of the interesting stuff is on YouTube, but that it is fairly high-effort because many of the videos are an hour long, or more. Maybe we’ll convince him to share some of his favorite videos soon in a future post!

A Few LibGuides

If you still need more, here are a few libguides:

What about you?

Who are your favorites to follow on social media? Are there helpful newsletters, blogs, podcasts, or anything else that we’ve missed? Let us know in the comments.