Find case law licenses

1. Indicate how many r&d licences have been (a) granted (b) refused

(a) 6 R&D licences have been granted, of which 3 have since expired.

(b) N/A – There isn’t a separate application route for R&D licences; there is a single application form for a transactional licence, and if the use is high risk but compliant with the MoJ’s principles, we grant an R&D licence. Therefore, there is no such thing as refusing an R&D licence; any refusals will have occurred prior to deciding whether a transactional or R&D licence would have been applicable. The number of refusals is therefore listed in answer 2b.

2. Indicate how many transactional licences have been (a) granted (b) refused

(a) 34 transactional licences have been granted, of which 1 has been terminated

(b) 1 application for a transactional licence was refused

3. Provide the public statements supplied by the successful licensees

Listed below are the public statements we have available for successful licensees. We do not have public statements for 15 licensees as they applied for licences before this was requested as part of the application process.

• BAILII
• Vizlegal
• ICLR
• Debt Squared Group Ltd.
• World Intellectual Property Organisation
• RELX Limited t/a LexisNexis
• Law Notion
• University of Oxford
• AI Artemida Technologies Ltd.
• Wexler AI Ltd.
• Moonlit B.V
• University of Edinburgh
• RELX Ltd. t/a LexisNexis
• LexMind
• LexisNexis BIS IP

James Irwin

On the application of contemporary LLMs to decisions for automated categorisation, metadata summary, and consequential indexation for real-time searching and analytics.
The service and analysis is Legal-tech in so far that it is aimed at legally trained persons for assistance in extended research of applied law.

Mishcon de Reya LLP

Advances in large language models, most notably OpenAI’s GPT-series of models, have unlocked a vault of possibilities extracting value from unstructured and semi-structured legal texts. However, it remains the case smaller NPL architectures, such as LSTMs, have a role to play in building lightweight, less costly solutions that are easier to test and introspect. In addition to conducting internal research on the use of LLMs with case law, MDR Research will use the data to train and open-source lightweight NLP model for data extraction and classification.

Doorda Ltd.

Doorda Ltd aggregates and links official public sources to enable easier access to data for compliance checks and risk mitigation. Court case information is one of many sources linked to registered UK companies.

David Quest KC

Current AI large language models are able to answer general legal questions but their effectiveness and accuracy is limited because they have not been trained on the text of judgments and current context limitations mean that a model cannot be prompted with a very large amount of text. Retrieval-augmented generation techniques are a potential solution, where the LLM can be provided with further context relevant to a particular enquiry, in the form of extracts taken from a database of published judgments. The project is to investigate the effectiveness and reliability of such a solution, and how it is affected by the parameters used for retrieval. The project is purely for research/experimental purposes, not to offer advice or for any other commercial purpose. The intention is to educate other in legal practice and academia about the possibilities and limitations of AI models for legal research.

Liga Zakon LLC

• creation of an informational and analytical product that will include judicial decisions that are available in the public domain on the resources of the National Archives;
• providing this product daily with new court decisions, as well as corrections or removals of previously published ones, based on data from the National Archives website;
• organization in the product variety of search possibilities for court decisions on:
• formal details and context (by the date of adoption, by the name of the court, by the name of the court decision, by the name of the court, by the names of the parties of the case, according to the words from the text, etc.);
• entities that are mentioned in the court decision (company, brand, person, address, CEO, normative legal act).
  We intend to process data to identify:
• the main participants in the process mentioned in the text of the court decision: the judge, the parties (plaintiff, defendant, prosecutor, third parties), representatives of the parties.
• other participants of the process (persons mentioned in the text of the decision, companies, brands), as well as related entities (addresses, CEO, norms of legal acts)
  The product will provide opportunities to search for court decisions on the specified groups.
• Extraction of necessary information (names, keywords, references to regulatory acts, legal precedents);
• Analysis of the extracted information for matches with similar information from other judicial decisions. For instance, if an individual is mentioned in different court rulings, the product user will be able to see this;
• Providing the user with the opportunity to explore publicly available information about the subject of the request, including legal precedents.

Naic Ltd.

Existing foundational models, such as ChatGPT, are so persuasive that one can easily be tricked into trusting them. The latest LLMs are based on the transformer architecture, which (significantly simplified) means that the model mathematically responds to users’ input with the most frequent or related sequence. However, in law, this logic is flawed, and the most frequent continuation is not necessarily correct.

Our project is an experiment aiming to use targeted and carefully chosen data, curated in a specific form and enriched with meta-tags, to mitigate the flaws of LLMs for legal tasks.

The plan is to fine-tune or train a large language model on carefully chosen and curated legal data and test it against existing foundational models to see to what degree the results are better.

This product is not meant for commercialisation yet and is a research project at this stage.

Byte Cognetics Ltd.

We seek to perform computational analysis of case law to integrate AI-driven legal research capabilities into Lawdec, our Case Management System designed for legal professionals. The purpose is to streamline legal research, making it more efficient and comprehensive, thereby enhancing decision-making and productivity within the legal sector.

This initiative primarily serves lawyers, legal researchers, and law firms by providing them with quick access to relevant case law and court judgments. The anticipated outcome is a more effective legal practice, with reduced research time and increased accuracy in legal arguments and case preparations.

Our methodologies involve using advanced natural language processing and machine learning algorithms to analyse vast amounts of legal text. This will allow for the extraction of key insights from case law, supporting more informed and strategic legal practices. We prioritise accuracy, impartiality and transparency and will review summaries to guarantee they provide reliable insights, upholding the integrity.

Tabled Technologies Ltd.

Libra empowers Litigants in Person with a Legal Information App and marketplace of Legal Providers, including pro bono legal clinics. Our innovative AI-driven chatbot will help users understand their legal issue, provide information and templates and refer users to dispute resolution services and legal providers for legal advice. The app will provide information and not legal advice.  Legal summaries of caselaw, generated by LLMs, may also help users better understand relevant legal situations.

A labelled corpus of the National Archives legal datasets will be leveraged by LLMs and NLP to efficiently analyse and summarise cases and legislation. We prioritise accuracy, impartiality and transparency. Our team of legal experts will review outputs.

We strive to democratise access to legal knowledge. We are committed to driving positive change in the English legal system, with a particular focus on litigants in person, by facilitating informed approaches to legal dispute resolution.

Wordsmith AI Ltd.

Wordsmith is a comprehensive AI toolkit designed specifically for in-house legal teams to enhance their document processing and analysis capabilities. Our platform is focused on streamlining legal workflows by automating routine tasks, which reduces manual effort and improves the quality and consistency of legal documents. This allows legal professionals to focus on more strategic activities, ultimately saving time and resources.

We cater to in-house legal teams, including lawyers, paralegals, and legal analysts, and extend our benefits to procurement and compliance departments within organisations. Our tools are instrumental in contract management and regulatory compliance, providing a robust solution for legal professionals.

Wordsmith employs AI-driven methodologies for document review and markup, data extraction, and summarisation. We offer custom AI assistants that are tailored to specific datasets, ensuring secure and efficient data handling without using customer data for training, thereby maintaining confidentiality and data integrity.

Lancaster University

I am a second-year PhD student at Lancaster University, UK, and I am currently researching applications of deep learning and natural language processing (NLP) in the legal domain. We conduct research on information retrieval, extraction and text generation using court cases.

We aim to develop methods based on language models (LMs) and NLP techniques to efficiently and effectively process court cases. Our main methodology for this research is using LMs, which have gained increasing attention in NLP; however, the legal domain has yet to benefit from them. As case law plays a huge role in the judicial process, the advancements of the technologies in this area would empower legal professionals with efficient tools.

As the outcomes, we will publish our findings in reputable and top NLP/ AI conferences (ACL, EMNLP, etc.) and journals (Language Resources and Evaluation, AI and Law, etc.) so that the community will benefit from them.

SQE Case Law Summariser Ltd.

The SQE Case Law Summariser Application helps aspiring solicitors enhance their understanding of the SQE syllabus by providing concise, accurate, and accessible summaries of case law. Using Retrieval-Augmented Generation (RAG) technology, it retrieves relevant sections from databases like FIND CASE LAW and simplifies complex legal language into clear, digestible insights. This tool supports law students and educators by aligning case law content with the SQE syllabus, enabling focused and practical learning. Key features include document retrieval, natural language processing for simplified summaries, and interactive feedback to refine content. The application fosters deeper comprehension of legal principles, empowering users to confidently engage with the syllabus and apply their knowledge effectively. By bridging the gap between complex legal records and practical learning needs, this innovative tool ensures a personalized and efficient study experience.

Leya AB

Purpose and Anticipated Outcomes:

Our initiative aims to enhance access to legal information and streamline legal processes through innovative technologies. The anticipated outcomes include improved legal research capabilities, optimised workflows for legal professionals, and more informed decision-making within the legal domain. By leveraging advanced computational analysis and AI (LLMs), our goal is to contribute to the efficiency and effectiveness of legal services.

Individuals or Communities Served:

This initiative primarily serves legal professionals and law firms, court users (including litigants in person), the Judiciary, and public bodies. By providing valuable insights and responses based on legal data, we seek to benefit these stakeholders and contribute to advancements in legal research and practice.

Methodologies or Activities Involved:

The project involves the utilisation of cutting-edge computational analysis (AI and LLMs) to process and analyse legal data. This includes the integration of legal documents, laws, and jurisprudence to generate accurate and relevant responses. The methodologies encompass continuous algorithmic refinement, adherence to ethical guidelines, and engagement with legal professionals to ensure the highest standards of quality and reliability.

108 AI Ltd.

108 AI is developing a legal analysis tool that leverages AI to provide precise legal insights, serving legal professionals and their clients by enhancing the efficiency and accuracy of legal research and advice. Our methodology involves processing historical legal documents with advanced AI techniques, ensuring improved outcomes in legal service delivery.

Neoflames Ltd.

At Neoflames, we are dedicated to innovating legal workflows through Wulo, our AI-powered platform tailored for conveyancers. Wulo harnesses the power of case law archives to simplify property transactions, ensure compliance with legal standards, and enhance decision-making for legal professionals.

Our platform is designed to serve conveyancers, legal practitioners, and property stakeholders, providing them with fast, accurate insights into legislation, case precedents, and local authority requirements.

Using advanced natural language processing and multi-agent systems, Wulo analyzes case law archives to extract critical information, identify legal risks, and deliver actionable summaries. At Neoflames, we prioritize data protection and compliance by implementing rigorous measures to safeguard personally identifiable information (PII) and adhere to all relevant data privacy regulations.

Through Wulo, Neoflames aims to empower the conveyancing community to streamline their workflows, mitigate risks, and deliver exceptional outcomes—all while maintaining the highest standards of security and privacy.

Blue J Legal Inc.

Blue J’s mission is to bring absolute clarity to tax law everywhere and on demand. By harnessing the power of generative AI, Blue J enables users to accomplish hours of tax research in seconds, delivering faster, more accurate answers and empowering tax experts to navigate grey areas with confidence, so that they can deliver meaningful insights for clients faster.

Blue J is designed to serve tax law specialists, including lawyers, accountants and other tax experts, seeking a robust, up-to-date resource for UK tax law to achieve better outcomes for clients.

We rely on retrieval augmented generation (RAG) and a curated database of trusted content, including HMRC guidance, and relevant legislation. Every answer is backed by authoritative sources, allowing users to easily verify findings at every stage. Our goal is to help experts save time so that they can provide superior client guidance and achieve better outcomes.

108 AI Ltd.

Casey is an AI-powered platform developed by 108 AI LTD, designed to enhance legal workflows by providing advanced contract analysis and risk assessment tools. By incorporating case law and legislative data from The National Archives, Casey aims to offer legal professionals comprehensive insights into complex agreements, facilitating informed decision-making. The platform utilises computational analysis to interpret legal texts, identify potential risks, and generate user-friendly summaries. Our goal is to support the legal community by delivering accurate, efficient, and accessible tools that uphold the principles of open justice and data integrity.

Faqtor

We are developing an AI-powered tool designed to provide free and publicly accessible question-answering and recommendation services related to UK law.

Project Overview:

Our project aims to democratise access to legal information and empower individuals to better understand their rights and responsibilities under UK law.  Many individuals face barriers to accessing legal advice, whether due to cost, geographical location, or complexity of legal language.

Our AI tool will address these barriers by:
Providing clear and concise answers to legal questions: Users will be able to ask questions in plain language and receive summaries of relevant case law and legal principles.

Generating personalised recommendations: Based on user-provided information, the tool will offer potential courses of action and suggest relevant resources.

Facilitating legal research: The tool will enable users to easily search and navigate through relevant case law, making legal research more efficient and accessible.

Braithwate Ltd.

RegPass is an AI-powered regulatory compliance platform designed to help financial services firms navigate legal and regulatory requirements efficiently. Our platform provides clear, actionable regulatory insights, enabling compliance officers, legal teams, and product leaders to stay ahead of evolving regulations while reducing operational burden.

We serve regulated financial institutions, compliance professionals, and legal teams who require precise and up-to-date regulatory guidance. Our goal is to enhance regulatory clarity, improve compliance decision-making, and streamline regulatory change management.

Using advanced Natural Language Processing and machine learning, RegPass analyses vast regulatory datasets, extracts relevant obligations, and delivers tailored insights based on a firm’s unique business profile. Our transparent audit trails, extensive filtering capabilities, and conversational AI interface ensure users can quickly access and understand the rules that apply to them.

We are committed to promoting regulatory transparency, reducing compliance complexity, and supporting financial services firms in maintaining adherence to legal requirements with confidence.

Peter Fields Legal

My mission is to transform the speed, accuracy, effectiveness and efficiency of dispute resolution. As a barrister, in addition to analysing established case law, I use my own AI tools to identify new cases in specific practice areas and assist in determining if/how they apply to cases that I deal with.

The efficiency and cost benefits of using AI in this way then flow to my instructing solicitors to enhance their service to lay clients and to reduce the costs of case management and dispute resolution.

My approach is captured in my book on effective dispute resolution (“A Practical Guide to Effective Dispute Resolution: Methods and Clauses, Dr Perova and Dr Fields, Sweet & Maxwell, 23.09.2024) including an in-depth analysis of how AI can be used in legal dispute resolution, current constraints and future opportunities.
I am regulated by the Bar Standards Board – Code of Conduct.

Rosa Esteban Muñoz

The PhD thesis will employ a corpus of case law to analyse the language present in judicial decisions. The study aims to identify linguistic patterns, trends and stylistic variations in a wide variety of legal decisions by using corpus linguistics methods. The research will focus on objective, evidence-based analysis, ensuring that case law is presented in its correct legal context, with proper citations. While the research is intended for a specialized audience, its findings are expected to be accessible and useful to a broader audience, including legal drafting experts and law students.

The anticipated outcomes will include a more profound, evidence-based understanding of the evolution of legal language, and the development of corpus-based analytical methods.

Therefore, the purpose of our study is linguistic and not related to the provision of legal advice. The data used will be presented faithfully, without distortion or selective editing that might mislead the target audiences.

Legal Utopia Ltd.

• Purpose and anticipated outcomes

The purpose of the “Find a Lawyer” (FAL) project is to develop a robust and comprehensive platform that assists individuals and businesses in locating qualified legal practitioners and entities. By leveraging data from The National Archives, we aim to enhance our system’s ability to provide accurate and reliable means to identify relevant providers by introducing a comparison framework. This comparison framework will help users selecting legal services providers. The anticipated outcomes include improved access to legal services, higher satisfaction rates among customers, and increased visibility for competent legal practitioners.

• Individuals or communities served

The project primarily serves individuals and businesses seeking legal assistance. This includes people facing legal challenges, corporate entities needing legal advice, and any other community members requiring professional legal services. Additionally, legal practitioners and firms benefit from the opportunity to showcase their qualifications and expertise through the platform as a business development tool.

• Methodologies or activities this will involve

The methodologies and activities involved in the project include collecting and integrating data from The National Archives and other relevant sources, developing sophisticated algorithms to calculate relevance and weight to overall metrics for legal professionals and entities, designing a user-friendly interface for efficient search functionalities, implementing rigorous quality assurance procedures, and engaging stakeholders for collaboration and feedback. These efforts aim to create a robust and accessible platform that enhances access to legal services, fosters informed decision-making among users, and promotes transparency within the legal profession.

4. Summarise what sort of audit, reporting or other process you use to monitor compliance

Summarise what sort of audit, reporting or other process you use to monitor compliance with the terms of the licences granted.]

The National Archives maintain a report of all licences issued, the type of licences issued and the applicable term (5 years for transactional and 1 year for R&D) and expiry date. When a licence nears its expiration dated, the Licensing department contact the licensee to establish whether an extension of the existing licence is required, whether the original parameters are unchanged and/or whether a change to the type of licence (i.e. R&D to a transactional) is applicable. If the later, the licensee is informed a new application form must be submitted for assessment. To ensure licence holders comply with the terms of the licences the Find Case Law editorial team maintain an up to date email circulation list of all licence holders so that they can inform them promptly of any judgments taken down from the service at the request of the court and otherwise answer any queries they may have about using our Atom Feed and API for pulling updated versions of judgments and newly published judgments.