“Artificial intelligence is not going to replace lawyers, but lawyers who do not use artificial intelligence will eventually be replaced by lawyers that do use it,” according to Stacey Cozewith and Matthew Eichen.
As published in the New Jersey Law Journal
June 26, 2024
What Is AI?
AI, or artificial intelligence, is an attempt by computers to, essentially, act more human. Generative AI takes text, image, audio, video, and code and generates new content. For example, I can ask ChatGPT to create a poem for my mother for Mother’s Day and the more details I provide, the better it sounds. Unlike traditional AI, which deals primarily with numbers and small amounts of text, generative AI can utilize all the modalities mentioned above and prepare entirely new content in any of the forms referenced.
Are We Using It Now?
As attorneys, we are utilizing AI every day. We use Excel to create spreadsheets. We utilize spellcheck on Microsoft Word and we utilize generative AI with Lexis Advance in our research tasks.
The growth of AI has been tremendous over the past few years. Five years ago, we were taking selfies with Snapchat, or utilizing other apps to transform our picture and aging us 80 years, making our voices 10 octaves higher, or to create a video of ourselves as dancing elves. Now, generative AI can create songs, movies, and text that look and feel as if they were created by humans.
What Are More Current Uses of AI and Future Possibilities?
Artificial intelligence is not going to replace lawyers, but lawyers who do not use artificial intelligence will eventually be replaced by lawyers that remain unfamiliar with this new technology. Artificial intelligence can and should be used in the workplace to increase productivity for attorneys and staff. ChatGPT is not the only artificial intelligence platform. In fact, many other commonly used programs already have their own version of artificial intelligence software incorporated into their database. Lexis AI+ was mentioned above. Another notable example is Adobe. This is a program most of us use every day. Adobe now has an AI add-on feature. We receive most documents and deposition transcripts in Adobe .pdf format. Now, with Adobe’s AI add-on, you can click a button and Adobe will summarize a deposition transcript, or other record/document you have uploaded.
If we focus some time into incorporating artificial intelligence into our daily practice, it will keep us ahead of the curve and free up time in the future.
An example of how we can utilize generative AI that we are not doing currently, is for discovery. You can upload documents to ChatGPT to review for specific information. For example, if I received a response from Amex from a subpoena that I served and they send 500 pages I may want to utilize an AI program to help me review them more efficiently. If I want to see if the other party is spending money overseas, or gambling, or sending large sums to family—I can upload the documents and ask ChatGPT to review the charges for any made outside of New Jersey, or in excess of $500 per charge. While it may not be perfect, the quick review of the generative AI program, can give an attorney a starting point or basis for a more efficient overview, while saving time and money.
Another example of use in discovery is if I want generative AI to compose a draft for discovery questions or forms. Let’s say I know the other party invested funds in crypto currency and I want to make sure my standard Interrogatories and Notice to Produce include questions on that subject. I can have ChatGPT draft appropriate questions for crypto discovery for my review. These AI generated questions may not be perfect, but again, they provide a good starting point for an attorney review and perfect.
Caution Is Warranted
We do have to be cautious about our use of artificial intelligence and what documents we upload to certain artificial intelligence platforms. Most importantly we have to be careful when dealing with privileged documents and medical records. One way to ensure that the documents we upload to AI platforms remain confidential is by utilizing a subscription service. For example, ChatGPT Plus is a subscription service. A ChatGPT Plus subscription gives you access to key features beyond general content creation, and most importantly for attorneys, the subscription includes confidentiality and security. With ChatGPT Plus, uploaded documents and data will not be shared with third-parties for marketing purposes as is the case with traditional ChatGPT.
Moreover, to be clear, any legal citations or advice or work product produced utilizing AI, must be fact checked. For example, a New York attorney was reprimanded for utilizing citations generated by AI without checking them. In the case of Roberto Mata v. Avianca, (Southern District of N.Y.) the defendant’s counsel wrote a letter on April 26, 2023 to the court questioning the authenticity of several cases cited by the plaintiff’s counsel in their opposition—namely, asserting that the cases did not exist. The court itself found that “[s]ix of the submitted cases appear to be bogus judicial decisions with bogus quotes and bogus internal citations.” In response, one of the plaintiff’s attorneys admitted that “[i]t was in consultation with the generative artificial intelligence website ChatGPT, that your affiant did locate and cite” the nonexistent cases.
Thus, we should not be relying on artificial intelligence completely, but we must start incorporating its use into our daily practice. Another example of caution in legal citations comes from Texas. A district court judge there made his own requirements regarding the use of artificial intelligence in court documents. Specifically, the judge orders all attorneys appearing before the court to file a certificate attesting that either: (1) No portion of any filing will be drafted by generative artificial intelligence; or (2) that any language drafted by generative artificial intelligence will be checked for accuracy by a human being. This precedent-setting Mandatory Certification is one of, if not the, first of its kind establishing the appropriate use of AI in legal proceedings—an issue lawyers are currently troubleshooting.
Evidentiary Issues With the Use of AI
Warren Buffet recently issued a warning that scams and AI powered attacks are extremely dangerous, and we have to be careful of letting “the genie out of the bottle” with AI. He said that his image and voice were recently replicated by an AI-backed tool, which was so convincing, it would have worked on his own family. Thus, we know that AI can be potentially utilized to fabricate photos, voice recordings, and other evidence.
A Notice to the Bar was issued in January of this year offering preliminary guidance on the use of AI by lawyers. The notice reiterated attorneys’ core ethical responsibilities under the Rules of Professional Conduct, specifically, diligence, confidentiality, and honesty (RPC 3.1, 3.3). But how can we be sure that evidence has or has not been produced or altered by AI? As stated, AI can be used to produce text, pictures, and videos. In fact, there are AI programs, which can intake a police report and churn out a video reconstructing the accident. Therefore, we must be prepared to address the issue of AI-generated evidence since as Warren Buffet warned, it can be a dangerous tool that is already here and used by many.
We are all familiar with video surveillance tactics utilized by insurance companies or the use of recorded conversations in a domestic violence matter. How can we be sure video surveillance of an injured client, a social media post, or a recorded conversation, has not been generated or altered with the use of AI? Issues such as this will require, at a minimum, careful scrutiny by the court. We must begin to tailor our discovery requests to specifically address disclosure of the use of AI in any aspect of the case. Further inquiry may be required into things like access to the algorithms and inputs, not just outputs; disclosure of quality assurance and systems integrity measures; assistance of experts; and so on.
This is an evolving area of the law and it is important that we are aware of the uses, advantages, and pitfalls of AI moving forward. These issues will be addressed on a case-by-case basis. We must be prepared to address the evidentiary issues and have forensic computer and other experts available to analyze the data to determine if it was falsely generated if the client insists the conversations, actions, or posts never occurred.
Conclusion
AI is a now commonly used tool that we must 1) be comfortable using and 2) ensure that all work product produced utilizing AI is fact checked and correct. While caution is always warranted, utilizing AI will assist our firm in becoming more efficient and will be a tool that most clients will expect us to use and expect that our bill reflects this use (such as not seeing extreme charges for document review that may be streamlined using AI).
We should ensure that the appropriate subscription services from ChatGPT or Adobe are provided to the attorneys/paralegals so that all documents uploaded are secure. We should use these tools in discovery and analyzing large tranches of documents and information such as several years of bank statements or deposition transcripts. We can use it to go through transcripts of trials for use in summations or appeals or to look for specific entries in discovery documents. While AI is far from perfect, it will assist us as attorneys in doing more higher-level legal work and not being marred down in document review. At its highest and best use, AI gives us a starting point for further review and analysis and will be utilized by most firms in the near future.
Stacey A. Cozewith is a partner at Sarno da Costa D’Aniello Maceri and concentrates her practice on matters related to family law.
Matthew R. Eichen is an associate in the firm’s personal injury department.
Reprinted with permission from the June 26, 2024 issue of the New Jersey Law Journal. © 2024 ALM Media Properties, LLC. Further duplication without permission is prohibited. All rights reserved.
