Friend or foe? When it comes to the legal profession, GPT-4 can be a bit of both – the Florida Bar

Friend or foe? When it comes to the legal profession, GPT-4 can be a bit of both

WhenGPT-4– Latest version Open AI’s Language model systems – wasIt was released in mid-Marchmany aspiring lawyers and law professors have used it to take it.Bar exam. ofA large language modelThe chatbot went through each subject andPerformed more than 90% of human tests.

This news was a bit of a shock, no doubt, and raised a number of questions.Preparing for the bar examFor most of the helpless people, he needs to study for 10 hours every day for about three months (after completing three years of law school). Suddenly, an artificial intelligence (AI) device can easily pass the bar.

What does this development mean for legal professionals? How will the development of language models positively or negatively impact the way lawyers learn, teach and practice in the new technological landscape?

Currently, there are ongoing discussions about how linguistic models can be useful.With the help of legal writing(as productionDrafting of original legal documents such as contractsAndConduct legal research). A junior attorney, for example, can save time by using GPT-4 to find relevant legal rules and regulations, identify potential conflicts in documents, and identify missed arguments. With this help, the lawyer can focus on higher-level tasks that require critical-thinking and analytical skills that language models cannot yet handle.

Compliance officers can use GPT-4 to create standardized document templates that ensure consistency in format, language, and structure. Because legal rules and regulations are often full of isocratic terms, GPT-4 is used to simplify language, clarify complex terms, and summarize long cases.

Language models can also help law enforcement officials conduct risk assessments. Officers can access information from internal sources such as employee emails and chat logs to identify compliance violations (such as fraud, corruption or other misconduct).

All these positive examples are certainly not failures. There are ethical concerns, and the breach of privacy is one of the issues with the language model among lawyers and law firms.

Some law and consulting firms have policiesWhere the tools limit or prohibit the use of language models, they may inadvertently reveal sensitive information and expose data breaches or cyber attacks. However, vendors can solve this problem by providing language model plug-ins that store the organization’s sensitive data in a proprietary database instead of sending it to language models or third parties.

Another issue is that legal errors created by language models can make lawyers responsible for the error. Clients trust their attorneys to work efficiently and diligently in their best interests. If a lawyer relies on a language model without thoroughly examining and verifying the resulting results, this can lead to false claims.

Lawyers must use language models carefully to perform tasks that require core legal knowledge. For example, when asked to write a legal essay, language models have shown that they are not yet fully capable of producing highly reliable legal analyzes (“yet” is the operative word here; the next generation is only a matter of time away. Language models outperform GPT-4).

Given the prevalence and impact of these language models, law professors must adjust their teaching methods accordingly. Teachers must develop an adaptive and innovative mindset by encouraging students to embrace new technologies. In the meantime, educators should emphasize the ethical implications and examine the appropriate use of these tools.

To better prepare law students for legal practice, educators need to identify and teach key skill sets such as legal rapid engineering, which involves properly preparing the inputs and outputs of AI tools to produce the right results.

Law schools can integrate AI courses into their current curriculum, providing them with the opportunity to experience language models and other AI tools for legal research, document drafting, and legal analysis. Courses leading to certificates using language modeling technology help students find employment after graduation.

Thus, while language models such as GPT-4 may present ethical and privacy-related challenges, they ultimately offer great opportunities for teaching and applying law. With care, lawyers can make these tools work to their advantage.

Jiaying Jiang, SJD is an Assistant Professor of Law at the University of Florida Levine College of Law. Her research focuses on policies and regulations related to emerging technologies, including artificial intelligence, fintech, blockchain, cryptocurrencies and central bank digital currencies. Professor Jiang engaged students in his fintech class in thoughtful and fruitful discussions on this topic.