Redefining Legal AI: Beyond A Case Law Database

Redefining Legal AI: Beyond A Case Law Database

In the world of legal tech, the integration of artificial intelligence (AI) has often been viewed through the lens of direct case law analysis and advice generation. A recent study by Stanford, however, highlights the limitations of this approach, revealing significant errors in legal information generated by large language models (LLMs) like ChatGPT (GPT 3.5, rather than GPT 4).

The Current State of Legal AI

The study underscores a crucial point: AI, in its current form, is not yet equipped to handle the complexities of legal judgment and case-specific advice reliably. The error rates, as high as 88%, raise legitimate concerns about the application of LLMs in legal practices. Asking out of the box AI to find case law and provide advice is foolish.

A New Approach to Legal AI Integration

Rather than directly asking AI for specific cases or legal advice, a more practical application lies in utilising AI for preliminary legal research. An example workflow involves using AI to generate relevant search terms for legal databases, streamlining the initial stages of case research.

The Proposed Workflow

  1. Receiving Client Requests: When a client request comes in, it is automatically forwarded to an AI system like GPT.
  2. Generating Search Terms: The AI analyses the request and provides targeted search terms relevant to the case.
  3. Database Query: These terms are used to query legal databases, automating the initial research process.
  4. Summarisation and Ranking: AI then summarises and ranks the returned cases, providing a brief overview of each relevant to the query.
  5. Lawyer's Analysis: Lawyers review these summaries and case rankings to make informed decisions, applying their legal expertise where AI falls short.

The Future of AI in Legal Practice

This approach to AI integration in legal workflows offers a balanced use of technology. It leverages the strengths of AI in data processing and information retrieval, while acknowledging the irreplaceable role of human expertise in legal analysis and decision-making.


AI, when used correctly, can significantly enhance the efficiency of legal research. However, it is essential to remember that AI is a tool to support, not replace, the expertise of legal professionals. The future of legal technology lies not in replacing lawyers with AI, but in harmonising AI capabilities with human judgement and expertise.