Use of Artificial Intelligence in Law

One of the greatest developments of recent times worldwide is undoubtedly artificial intelligence. Especially ChatGPT, which was released by OpenAI on November 30, 2022, has managed to attract the attention of millions of people in a very short time with its capabilities, and with GPT-4, which was released on March 14, 2023, it has shown that there is perhaps no end to the work that artificial intelligence systems can do and the points they can reach.

GPT-4 has entered our lives with many innovations. It can continue a poem or song written long ago in the same way, evaluate memes and understand the humour content, even explain why that image is funny, write code for the application, find errors in written code and design a website as described. In addition to these, GPT-4 has been tried in many exams. While artificial intelligence had a lower percentage of 90% in previous versions, it increased the success rate to the top 10% with GPT-4 in the Uniform Bar Exams, which is one of them. In short, artificial intelligence can now do things that were unpredictable in the past very simply.

The recent development of artificial intelligence has affected the field of law as well as every field. It is currently debated whether the artificial intelligence program DoNotPay, the first robot lawyer founded by Joshua Browder, a 19-year-old college student in 2015, can go to court. DoNotPay, which was designed to help its users against simple transactions such as parking tickets when it was first established, now supports its users in many legal activities, such as registering a trademark, filing a lawsuit and obtaining a divorce certificate. DoNotPay was expected to attend his first hearing on February 22 and defend his client in a traffic ticket case. However, Joshua Browder said that he had received threats from State Bar Prosecutors and that if DoNotPay joined to trial, he could face six months in prison. Therefore artificial intelligence could not take part in the hearing on February 22. Even though a robot lawyer was prevented from appearing in court, DoNotPay continues to help its users through its website.

With the recent development of artificial intelligence, it is seen that it is more successful than 90% of people in the Uniform Bar Exams and can do simple legal transactions for now. Today, artificial intelligence is also used as an expert. Thanks to the rapidly developing artificial intelligence programs, it may also appear in cases as a judge or prosecutor in the near future.

On the one hand, The use of artificial intelligence in the courts has the potential to provide considerable benefits in the law. First of all, there are too many cases in the courts nowadays, and most cases are not resolved quickly enough. With artificial intelligence, especially simple cases may be resolved swiftly, reducing the density of cases in the courts and protecting the right to be tried in a reasonable time. In addition, it may be prevented from making different decisions for similar situations with the decisions that artificial intelligence can make; thus, the public's confidence in the law can increase. Finally, It can be said that artificial intelligence can provide impartiality better than judges when making decisions because it does not have emotions and preferences.

On the other hand, there are also potential risks to using artificial intelligence in court. Firstly, the judge has the discretionary power specified in Article 4 of the Turkish Civil Code. According to this article, the judge should consider equity while using his discretion. It is not known whether artificial intelligence has an idea of equity. In addition, it is unclear what criteria artificial intelligence can calculate non-pecuniary damage claims. Assuming that artificial intelligence, which enables it to make decisions that do not conflict with each other in law, is used in all legal remedies, there will be no difference in the way artificial intelligence evaluates the issue, so legal remedies will have no legal value. Finally, it is doubtful whether artificial intelligence programs, which global companies own, can fully protect the personal datas of the claimant and the defendant, which will be added as input to them when used in courts.

As mentioned above, using artificial intelligence in litigation has many advantages and disadvantages. However, in my opinion, the use of artificial intelligence in law should not be decided as "it should be used completely" or "it should not be used at all". Although it is not yet possible for a trial to be conducted entirely by artificial intelligence due to the potential risks, it can make a significant contribution in helping judges, prosecutors and lawyers to reduce the current caseload and provide better legal services thanks to the work and skills it can perform.

Finally, I asked Artificial Intelligence via ChatGPT for its views on the use of artificial intelligence in law. Here is AI's answer on this subject;

Question: The use of artificial intelligence in court cases has some risks, such as the protection of personal data, as well as certain advantages, such as enabling quick decisions. Do you think artificial intelligence should be used in law? If you think it should be used, what size should it be used?

Artificial Intelligence: As an AI language model, I do not have personal beliefs or opinions, but I can provide an objective perspective on the use of artificial intelligence in law.

Artificial intelligence can potentially bring several benefits to the legal system, such as automating repetitive tasks, analyzing large amounts of data, and detecting patterns that might not be apparent to human observers. This could lead to more efficient and accurate decisions, reducing the workload of judges and legal professionals, and enabling quick decisions.

However, the use of AI in the legal system also poses risks and challenges, such as protecting the privacy of individuals, avoiding biases and discrimination, and ensuring accountability and transparency in decision-making processes. For instance, AI models may be trained on biased or incomplete data, leading to unfair outcomes and reinforcing existing inequalities.

Therefore, the use of AI in the legal system must be carefully regulated and monitored to ensure that it complies with ethical and legal principles, protects the rights and interests of individuals, and promotes fairness and justice. AI should not replace human judgment, but rather complement it, by providing additional insights and recommendations that can be reviewed and evaluated by legal professionals.

The specific areas where AI could be used in the legal system will depend on the specific context and jurisdiction, but some potential applications include document analysis, case management, legal research, and prediction of legal outcomes. The size and scope of AI deployment will depend on the specific needs and capabilities of each organization and jurisdiction but should be subject to ongoing evaluation and assessment to ensure that it delivers the intended benefits while minimizing the risks and unintended consequences.

Bibliography

• https://www.cbsnews.com/news/robot-lawyer-wont-argue-court-jail-threats-do-not-pay/ (Last Accessed: March 25, 2023)
• https://gizmodo.com/donotpay-robot-lawyer-ai-parking-ticket-1850031456 (Last Accessed: March 25, 2023)
• GPT-4 Technical Report
• An International Look On The Use Of Artificial Intelligence In Court Decisions And Opinions About Robot Judges, Hikmet Bilgin, 2022

---

Tagged with: Technology, Artificial Intelligence, AI, ChatGPT, OpenAI, GPT-4, DoNotPay