On the Ground Report: While AI Simplifies and Saves Time, Lawyers Still Taking the Lead with Hearts and Minds

This article was first published on 15 December 2024 in the Singapore Mandarin broadsheet, Lianhe Zaobao.

SLW obtained permission to reproduce the article to give the legal community a broader view of legal reports for various news syndicates.

Artificial intelligence is making inroads across various sectors; the legal industry is one of them. Not only do lawyers have to master new technology, they also need to reassess their professional role. This week’s On the Ground Report pays a visit to law firms to understand how technology helps lawyers handle cases and explore how AI curriculum is being incorporated into legal education in Singapore in order to cultivate an innovative mindset in new generations of legal talent. The Singapore Academy of Law (SAL) also shares how the platform for assessing legal material is being optimised using generative AI, to bring about greater convenience for legal practitioners.

Rajesh Sreenivasan, 55, partner at Rajah & Tann is one of more than 6,000 lawyers in Singapore and he is as an advocate of AI. He joined Rajah & Tann in 1998 and currently heads a team of more than 50 lawyers from 10 countries, handling cases related to technology and data. The clients they serve include some internet giants.

If a client company experiences a cyber attack, Sreenivasam’s team has to instruct the client to make a report to the Personal Data Protection Commission (PDPC) within 72 hours, in accordance with local laws. Following that, the team will have to comb through volumes of legal text, meet with the clients, launch a professional technical investigation, and finally, come up with a solution.

In recent years, incidents of data breaches have been on the rise in Singapore. According to figures from the Ministry of Digital Development and Information, there were more than 200 data breaches in the public sector in 2023, an increase of 10% from 2022.

From his office at Marina One, Rajesh Sreenivasan walked us through a work day in the life of a lawyer using a simulation of a data breach incident.

He opened a generative AI legal tool, and through a platform similar to ChatGPT, asked, “Are there any laws and regulations in Singapore pertaining to data portability?”

AI’s response: Data portability has already been introduced into the Personal Data Protection Act amendment.

“Lawyers need to be cautious with these machine responses, if they are not careful it could result in a huge mistake, and the best way to ensure accuracy is to keep asking follow up questions,” Sreenivasan said, pointing out the weaknesses of machines while continuing with his line of questioning. “Has data portability liability actually been implemented in Singapore?”

This time, the machine said: data portability and related liabilities are not currently in effect in Singapore.

Sreenivasan explained: “Just now the machine used the verb introduced, which is misleading, ‘introduced’ and ‘used’ are two different matters. When we come across ambiguities like that, the lawyer has to take further steps to verify the information.”

Generative AI tools require human verification but cut time spent by 30%

Although generative AI tools require human supervision, their use can enable time savings of 20 to 30% overall, said Rajesh Sreenivasan. In addition, he also uses Microsoft’s Copilot to record video conferences, generate meeting minutes, summarise presentations and plan meetings. Copilot can also help lawyers save 30% of their time.

“The time saved allow lawyers to better interact with their clients, and complete more complex legal work,” he said, citing divorce and other civil cases as examples that require lawyers to spend a lot of time communicating with the parties involved. “The more the enquiries are fine-tuned, the more details can be uncovered; in many cases, lawyers can even save marriages.”

“Even though divorce is not my specialty, but I know that when an outstanding lawyer receives divorce enquiries, their first reaction is never to ask ‘when’, but rather, ‘why?’”

Aside from technology and civil cases, AI can also support financial cases. Meritus Law LLC is a firm with 10 lawyers specialising in dispute resolution, arbitration, mergers and acquisitions, financing, and other legal services.

Alfred Lim, the firm’s managing director, said it uses Zoom and Perplexity, which are both software equipped with AI functionality. Zoom can produce meeting minutes that are 75% accurate, which are then edited manually. Perplexity, despite not being designed especially for the legal industry, is able to help lawyers search for online materials very efficiently.

Lim believes AI performs very well in summing up the latest legal perspectives across different jurisdictions, and the latest sentencing norms in criminal cases. The firm is currently planning to provide dedicated training on the use of AI for its employees. He said, “In future, we would still require professionals with legal training to check and verify machine-generated documents in order to ensure accuracy.”

Eric Chin, Asia Pacfic director of NewLaw PwC, said the company uses Harvey, Leah by ContractPodAI and other generative AI tools to provide clients with solutions that are human-led and tech-driven.

According to Chin, the most significant transformation AI has brought can be seen in how AI tools are enhancing the capability of professionals. “After the AI generates a draft text, our experts will check the copy to ensure stringent quality control,” he explained. “This enables us to give full play to the advantage of combining professional expertise and AI.”

To meet the needs of the legal profession, businesses around the world are also competing to develop and promote new AI tools for lawyers. Thomson Reuters, the international information services provider headquartered in Canada, invests more than US$200 million (S$268 million) in AI research and development each year. The company has launched CoCounsel, which it claims to be the world’s first professional grade generative AI assistant, which can aid legal practitioners in the preparation of statements, drafting letters, database search, document summaries, extracting contract data, and ensuring contracts compliance and policy compliance, amongst other functions.

Lee Hui Li, managing director of Microsoft Singapore, said many law firms in Singapore are using Copilot – which comes with Microsoft 365 – as their daily AI assistant. From managing their mail boxes to drafting documents, from designing client presentations to using Teams, lawyers have been able to raise their efficiency with the tool.

Agencies work together on AI platforms to complete basic tasks for lawyers

Singapore has taken the lead in setting up an AI platform for legal practitioners, enabling local lawyers to enjoy a technological advantage in dealing with cases.

Established in 1988, the Singapore Academy of Law (SAL) currently has around 13,500 members, including lawyers, law professors, judges, public prosecutors, and law school students. The organisation aims to drive transformation in the legal profession, raising the technology capability of lawyers through multiple training programmes, and aiding law firms in their digital transformation. LawNet, which is managed by the Academy, brings convenience to lawyers in their search for judgements, detailed legal amendments, and other services.

Yeong Zee Kin, chief executive of the Singapore Academy of Law, 52, chose to take up the only legal computing class available when he was reading law at the National University of Singapore. For his masters degree, he also opted to specialise in computer and communications law. Speaking to Lianhe Zaobao, he said more than more local lawyers are using AI tools in their daily case work, and SAL was working closely with the Infocomm Media Development Authority (IMDA), AI Singapore, and other agencies to build AI platforms for them.

“Lawyers are still playing a leading role in case work, but AI tools can take their place in some more basic tasks,” Yeong said.

He added, many young lawyers use ChatGPT and other generative AI tools in their work, and some also use Harvey and other professional AI tools. At present, lawyers are using mainly the summary and conclusion functions of these tools.

An important part of a lawyer’s work is reading legal material, and this is usually split into three steps: locating and identifying the material, reading, and then providing an opinion based on the material.

“AI can help lawyers become more efficient in identifying information, but they still need to do in-depth reading and comprehension, and analyse laws and facts, in order to give useful legal advice and opinion,” Yeong said. “After they have completed the initial draft of their opinion, AI can help to edit the draft.”

Yeong said through SAL’s collaboration with IMDA to make use of AI technology, more than 15,000 abstracts from court judgements have been produced and made available to lawyers through the new LawNet. Since 1965, Singapore courts have published 25,000 judgements, but because of a manpower shortage, only 9,000 have been included in the Singapore Law Report.

Using AI to create abstracts free lawyers up to consider strategies and analyse evidence

Judgments are final decisions made after court trials, and their lengths vary with the complexity of the case. The judgements on straightforward and short cases may only run to a few pages, but for more complicated or important cases, judgements can be a few hundred pages in length. Abstract-generation help lawyers save time so that more time and effort can be put into analysing the law and the evidence, formulating legal strategies, and other more important tasks.

Another task that requires a lot of hours are the reading of statutes. Lawyers can only read up professional literature to understand how the statues have been interpreted across different cases, and they are also limited by the author’s area of specialty. To expand the legal horizon, the Academy is working with AI Singapore to provide more interpretations on the LawNet platform, to help lawyers understand the different legal perspectives and improve their work.

Even though setting up an AI platform for legal material helps to improve efficiency for legal practitioners, Yeong pointed out that as the barriers for AI tools become lower, clients can also easily use AI to generate documents. This will put on more exacting demands on lawyers as they need to a pair of more discerning eyes to tell apart true and false, and also the ability to truly master technology.

Do what AI can’t do to create value  

Experts from all three law schools in local universities are of the view that lawyers will continue to have a job and will not be so easily replaced, as long as they continue learning about AI and mastering new technology.

Andrew Simester, dean of NUS Law, pointed out that AI will completely transform the traditional service model of legal consultation, with professional AI generative tools like Harvey greatly increasing the efficiency of tasks such as drafting documents, reviewing contracts, abstracting documents, and client communication. But these tools still require human supervision.

Leslie Chew, dean of the School of Law at Singapore University of Social Sciences  (SUSS), believes that while generative AI has replaced or will be replacing lawyers in performing some routine tasks, not all of a lawyer’s work can be taken over. “There’s no simple equation for the impact the prevalence of generative AI on the number of lawyers needed.”

Chew gave an example: in an M&A deal, the first draft agreement can be created using AI more efficiently than a junior lawyer. But the role of junior lawyers can be redefined, and young lawyers should be aware that they must be able to do what AI cannot do in order to create true value.

He explained that AI can only be used to compile information at the moment, and is still unable to carry out practical analysis, “In the legal field, generative AI can reference relevant laws and cases, but is unable to apply the law to practical situations, nor provide a reasonable solution.” But he conceded that generative AI will be able to make advances in application in the future.

On the career development of new lawyers in the age of AI, Lim How Khang, assistant professor of law and computer science at the Singapore Management University, is of the view that new lawyers will have to learn and master work processes that are based on AI, focus on reviewing content produced by AI, and take on more important analytical and coordination tasks. As the use of AI spreads in the legal profession, lawyers have the responsibility of using the tools to perform their duties to their clients, the courts, and the judicial systems better.

Lim observed that the adoption of AI across the legal profession is uneven: large international firms that have ample budget and business volume will be the first to use AI tools; while smaller firms and the legal department of government agencies may proceed with more caution, creating tech solutions according to their specific needs.

Lawyers should focus on honing their judgement, intuition, creativity and other complex tasks

Simester said as more legal firms bring in AI tools, graduates of NUS Law should also have a firm grasp on their use. To this end, the law school has started compulsory law and technology classes in which undergraduates learn basic programming and computing principles. Technological innovation has also given rise to new legal research areas, such as legal issues in the use of copyrighted work in AI. NUS Law is also expanding elective curriculum on data protection, the regulation of digital platforms, and other related domains, and offering students opportunities for cross-disciplinary learning.

Lim How Khang said the SMU Yong Pung How School of Law has overhauled its curriculum, and students can learn the basics of the intersection of law and technology through courses such as Law and Technology and Digital Intelligence for Lawyers. In the advanced courses, students are given the opportunity to take courses on legal-related large language models and generative AI, so that they can gain more in-depth knowledge on the use of AI and language models.

Leslie Chew of SUSS said the university has also established a standard protocol for the use of generative AI. The school is not against the use of generative AI by students; instead it offers them guidance on how to use AI responsibly and efficiently. Today’s undergraduates should realise that they are entering an age where information is readily available. In this environment, knowledge itself no longer offers any added value and may not be considered an advantage in the professional world. “The real test is creative thinking and the ability to apply knowledge to solve real life problems – this is especially important for lawyers and other professionals.”

Source: Lianhe Zaobao © SPH Media Limited. Permission required for reproduction.

Third executive fired by SingPost to contest decision, says move ‘unfair and without merit’