Generative AI has made a significant impact on how we work since Chat GPT launched in 2022. Large Language Models (LLMs) like ChatGPT have evolved rapidly, improving human-machine interaction. For lawyers seeking a gen AI powered assistant, it’s best practice to consider the risks and benefits.
Singapore Courts intend to enhance access to justice with generative AI tools. Justice Aedit Abdullah, Judge in charge of Transformation and Innovation explained how Singapore Courts is exploring the use of generative AI solutions to ‘facilitate access to justice for court users.’
“There is a desire for simpler, faster procedures, saving time and resources… Technology promises the possibility of meeting these needs,” said Justice Abdullah.
Speaking at Singapore University of Social Sciences (SUSS) engagement series ‘Conversations with the Community’ in May, Justice Abdullah outlined Singapore Courts’ responsibility to uphold a justice system that is ‘accessible and meets everyone’s needs.’
“The courts have an obligation to assist the self-represented person,” Justice Abdullah added.
The Small Claims Tribunal (SCT) addresses minor civil disputes in Singapore without lawyers. It is an efficient and cost-effective alternative option for individuals or small businesses. The SCT hears about 10,000 cases each year. It has now become the testing ground for Singapore Courts to test the use case for generative AI system deployment and implementation.
“At least in the provision of legal information, so that the self-represented person can consider what steps can be taken in his or her own interest.” Justice Abdullah concluded.
How can lawyers in Singapore use gen AI at work?
Increased productivity and efficiency are key factors that influenced widespread AI adoption. Professionals in Singapore want to invest their time in high-level work according to our recent survey, and would gladly delegate admin to a gen AI assistant, if given the chance.
According to the latest Future of Professionals Report – Asia & Emerging Markets Edition, professionals want to deploy gen AI for:
- Work product creation: Drafting, refining, or editing documents, and analysis.
- Reduction in admin: Unlocking time for focused high-level professional work.
- Intelligent interfaces: Conversational/assistive software.
- Intelligent information retrieval: Searching for and interpretation of information across databases and sources.
For lawyers seeking a gen AI powered assistant, it is important to weigh up the risks and benefits. One of the major pitfalls that busy professionals can fall into is over reliance on what generative AI systems can deliver. The infamous Mata v Avianica case in the US demonstrates human oversight is absolutely mandatory when working with generative AI.
How does the RAG model help?
Retrieval Augmented Generation (RAG) is a framework that can limit where an AI system searches for information. The RAG acts as an AI guardrail, focusing the generative AI system it works with to search and retrieve information solely from verified sources.
Thomson Reuters’ CoCounsel, designed for lawyers, is based on the RAG model. CoCounsel operates within trusted, secure guardrails and draws information solely from verified data sources. Since CoCounsel launched last year, 45 large US law firms that employ more than 50,000 lawyers have rolled out the generative AI system.
Most recently, CoCounsel launched in Australia and Canada, and now Singapore. The launch was a milestone for customers according to Jackie Rhodes, Managing Director, Thomson Reuters Asia & Emerging Markets.
“Now this market can take advantage of the power of gen AI with the data and security guardrails necessary to meet professional legal standards,” said Jackie.
In 2023, Thomson Reuters acquired CaseText Inc. Founded in 2013, Casetext has leveraged machine learning and Large Language Models (LLMs) to create generative AI solutions for legal professionals. According to Jake Heller, Head of Product, CoCounsel, Thomson Reuters bringing CoCounsel to markets that, until now, haven’t had access to professional-grade AI is a tremendous accomplishment.
“As we empower more customers to realise tremendous improvements in productivity and work product quality, we’re motivated to move even faster to bring the power of generative AI to more of the world,” said Jake.
What is CoCounsel and what can it do?
CoCounsel is Thomson Reuters’ generative AI assistant. Designed to maximise time for lawyers by making quick work of routine tasks with functionality across:
- Intelligent document analysis
- Sophisticated document search
- Summarise
- Timeline
- Contract policy compliance
According to David Wong, Thomson Reuters’ chief product officer, CoCounsel can deliver answers enabling lawyers to focus their time, energy, and expertise on meaningful, more impactful work.
“Our unique combination of resources means we can deliver on our vision of providing professionals a new, human-centric point of access to our suite of Thomson Reuters products,” said David.
“As our products continue to expand and improve, the customer experience will remain centered in our GenAI assistant as CoCounsel learns new skills and capabilities – unlocking productivity and becoming the way professionals work.”
Core CoCounsel skills designed specifically for lawyers
There are eight essential CoCounsel skills. Learning how to leverage the skills available can maximise time for lawyers to do impactful work.
- Search a Database: CoCounsel will search through documents like deposition transcripts and compiled litigation records from your law firm’s database. The genAI system will prepare a memo with a summary and bookmarks pointing to the sources it retrieved the information from. CoCounsel will provide the excerpts and hyperlink to the documents referenced. Users can ‘search by a concept,’ which is a level-up from a basic keyword search on traditional platforms.
- Review Documents: CoCounsel can look for information in documents to answer queries. Our genAI system will specify pages it sourced the information from within the documents it interrogated.
- Contract Policy Compliance: Feed CoCounsel your contract playbook or ask it to create one based on a standard precedent. Use this document to analyse internal or third-party contracts against your internal position with automated amendment redlining directly in Microsoft Word. CoCounsel will explain how it arrived at its answer, referencing the relevant section in your contract playbook, so that lawyers can fact check the work it has produced.
- Extract Contract Data: CoCounsel can find specific information within contracts. The gen AI system will produce a memo with the answers required. It will also include hyperlinks to each part of the contract the answers were located.
- Summarise: CoCounsel can produce an overall explanation after interrogating large, complicated documents, or a set of documents. It is important to note, focusing on particular aspects, angles or lines of questioning, is for ‘Review Documents’ as it is designed to be far more granular.
- Prepare for a deposition: CoCounsel can produce an initial outline of topics and questions for a witness in a deposition, or a pre-trial witness examination as it is known here in Australia. The genAI system can also recommend topics and questions that you can include in your deposition outline. The gen AI system breaks down a set of facts into individual issues. It will then create a list of questions under those issues. This functionality also works well for creating an affidavit or a client intake.
- Timeline: CoCounsel creates a chronology of events based on a set of documents. The gen AI system is capable of reading handwriting. The tool is designed to OCR scan to seek information within scanned documents. Timeline can go through large sets of documents and pull out all the important events.
- Draft correspondence: CoCounsel supports lawyers by drafting correspondence efficiently. Our gen AI model will draft letters, emails, memos and the like after reading the description of the correspondence of your request.
As AI skills in legal practice become more commonplace, a new way to describe AI is emerging. Learn about the concept of ambient AI and what it means for the future of lawyers.
