Artificial Intelligence (AI) is a term used to describe technology that allows machines to perform tasks usually requiring human involvement. We already use AI daily - think Siri, Alexa, and even predictive text! But now, many in-house legal teams have embraced AI to optimise workflows and free their time for higher-value tasks.
These AI Assistants, powered by analytics and cognitive computing, are revolutionising the legal profession. With their ability to mimic human responses through natural language processing and large language models (LLM), they are not just a buzzword. They provide a practical solution to boost productivity and enhance efficiency in in-house legal teams and automate the time-consuming, routine aspects of contract lifecycle management (CLM), such as document processing, sorting and summarising files, contract review, and compliance.
Let's take a look at five of the best AI Assistants for in-house legal teams. Each assistant brings unique features and benefits, catering to different business requirements, and this overview will provide you with a great idea of how they could work for your organisation.
Summize
Summize, a CLM tool powered by Summize AI, offers a simplified way for legal and commercial teams to manage digital contracts across the contract lifecycle. Its AI Assistant in Microsoft Word and Word Online brings speed and automation to any business, as well as intelligence to contract creation and management. Summize's tool transforms long and complex documents into easy-to-read, digestible summaries. Plus, the ability to directly edit and review contracts using AI technologies in Microsoft Word not only improves efficiency and improves business-wide adoption, but also empowers in-house legal teams, mitigating risk, and streamlining the contract review process.
Summize easily connects with tools such as MS Teams, Slack, Salesforce, Outlook and more! It leverages Azure's global top-tier security to ensure customer data is safe and secure and uses the Summize AI Architecture to avoid cross-sharing of information.
OpenAI (ChatGTP)
ChatGTP, a customisable generative AI, is a powerful tool for understanding and generating natural language, making it particularly valuable for tasks like contract creation. OpenAI encourages users to customise their own AI programmes using a ChatGTP model, which empowers in-house legal teams to adapt the AI to their specific needs and streamline and automate legal processes. In seconds, it can generate robust, professional contracts and clauses or identify legal risk.
However, ChatGTP requires careful and planned input to avoid getting the wrong outcome - prompts must be unambiguous and generalisations must be avoided. Lawyers will also need to review the output carefully to ensure accuracy and prevent mistakes and errors produced by ChatGTP. There’s also the risk of breaching confidentiality if the data provided to train the AI is shared with others.
Additionally, ChatGTP is based on LLMs, but these have an expiry date – data is limited up to 2021, unless you purchase an upgraded package, so it won’t be able to keep up to date with new legislation.
Microsoft CoPilot
Microsoft CoPilot is an AI-powered tool designed to help Microsoft 365 users improve communication and boost productivity by automating repetitive tasks, such as summarising documents. It is a sophisticated chatbot based on LLM; users ask questions in a conversational style, and CoPilot generates answers in real-time based on documents, emails, calendars, chats, meetings, contacts, and other business data.
However, context is vital, and questions require as many keywords as possible, requiring human interaction and effort. Microsoft Co-Pilot is also only available for the Microsoft 365 suite with an upgraded licence and it can’t be integrated with other tools and software at this stage.
Thomson Reuters – Document Intelligence
Based on Microsoft Word, Document Intelligence from Thomson Reuter uses AI trained by experienced Practical Law editors to read and organise documents and manage workflows. It can expedite drafting and improve negotiations, saving time and money on contract review and data extraction.
However, users suggest that additional training sessions or how-to guides are required to use the tool, as the functionality and UI is not particularly straightforward, and it has also been mentioned in other feedback that the summaries are not completely accurate.
Casetext
Casetext's essential product, CoCounsel, is an AI legal assistant powered by GPT-4 that performs document review, deposition preparation, contract analysis, and timeline creation within minutes. It identifies critical documents in contract databases and automates contract review.
On the whole, However, users say the search is not intuitive and requires too much data to get the correct result. Other customers say there is a lack of customisation and currently an inability to access briefs from all cases. The tool was also originally created for law firms and solo legal practitioners, so in-house lawyers may not find it as useful as other more tailor-made AI Assistants.
The Drawbacks of AI Assistants
All AI Assistants will face some limitations as AI is not a perfect solution yet. For example, most tools require their technology to be trained - it must ‘learn’ from a dataset of hundreds, even thousands, of documents before it can conduct a contract review entirely on its own. It must also be backed up by expert human knowledge.
The dataset must also be rich and accurate: AI Assistants using ChatGTP often get their information from scanning websites, trusting what they ‘see’ as truth, making it vulnerable to misinformation and potential copyright infringement. AI can also introduce automation loops that increase risk or cause unnecessary back-and-forth; for example, some tools are so heavy on redlining that they may miss commercial decisions or trade-offs an in-house lawyer would spot.
The AI Assistant you choose will depend on your needs; its features must solve the problems you need to resolve, be easy to use and integrate into your daily working processes.