Artificial intelligence is here to stay in the legal sector. More and more lawyers, legal executives and legal departments are adopting AI tools to optimise their work, from document review to process automation. However, to get accurate and useful answers from AI, knowing how to interact with it is critical. That's where legal prompting comes in.
If you have used artificial intelligence (AI) for your legal tasks and feel that the results are not as accurate as they should be, or if you have not yet explored this technology and are wondering how to leverage it in your day-to-day work, this article is for you. We dive into legal prompting, how it works and how to create effective prompts that optimise your legal practice. In addition, we will explore case studies and strategies to improve your interaction with AI, so that you can turn it into a true ally in your daily work.
This article is available in Spanish.
Table of contents
Legal prompting is the technique of formulating precise instructions to a generative AI system to obtain relevant and adapted answers to the legal context. It involves strategically designing questions or prompts so that the AI understands the legal framework, key terms and the specific purpose of the query, in order to obtain the answers best suited to the user's needs.
In the legal sector, where precision and detail are essential, a good prompt can make the difference between a superficial and empty answer and a truly useful one for the lawyer. As legal professionals adopt AI to review contracts, analyse regulations or generate documents, knowing how to structure a prompt correctly has become an essential skill.
Legal prompting enables lawyers and legal professionals to optimise the use of AI in multiple areas of their work. Some key applications include:
Lawyers can use AI to draft contracts or review clauses. A good prompt can be used to request a specific clause with certain regulatory requirements or ask for a contractual risk analysis, as well as obtain legal opinions in a matter of seconds.
Instead of reviewing lengthy legal texts, a well-formulated prompt can ask the AI to extract key points from a judgment or contract and compare them with existing regulations.
Companies can use legal prompting to check whether their contracts and policies comply with regulations such as the General Data Protection Regulation (GDPR), and local or international regulations.
In-house legal departments can train AI systems with specific prompts to answer frequently asked questions from other areas of the company about regulations, compliance or internal processes.
A prompt is the instruction provided to a generative AI to generate a response. In the legal domain, the quality of the prompt with which the AI interacts, usually in a text or audio format, determines the relevance and accuracy of the information obtained.
There are different types of prompts, including:
To obtain quality answers in the legal field, prompts must be clear, specific and well-structured. This requires daily experimentation with AI, constant trial and error, but also specific technical training in legal prompting. But to get you started, here are some basic steps you can follow:
Before writing, ask yourself: do you need an informative answer, a comparative analysis, a legal opinion, or a legal document?
Instead of asking ‘How does data protection work?’, better formulate ‘Explain how the GDPR regulates the transfer of data to third countries with examples.’
AI works best when it has context. A prompt such as ‘write a service provision contract for a consultancy in Spain’ will generate better results than simply asking for ‘A service provision contract.’
Tell the AI how you want to receive the response:
‘Summarise the following judgment in 5 main points and provide an analysis of its impact on labour law.’
You may be interested in: Guide to optimising your legal prompts for generative AI
Here is a concrete example of a well-formulated prompt:
‘Draft a bilateral non-disclosure agreement (NDA) between two companies entering into a business relationship, using precise, technical language aligned with legal industry best practices. The agreement should contain essential clauses, including:
(i) the subject matter of the NDA and the detailed definition of ‘Confidential Information’, ensuring the inclusion of commercial, financial, strategic and technological data;
(ii) the obligations of both parties regarding the protection, use and disclosure of the information;
(iii) the duration of the agreement, specifying both the term of the NDA and the period thereafter during which the information must remain confidential;
(iv) exceptions to confidentiality, indicating cases in which information may be disclosed by legal mandate or express consent;
and (v) dispute resolution mechanisms in the event of a dispute.
In addition, incorporate a specific penalty clause for breach, establishing a penalty of 10% of the value of the underlying contract or, failing that, of the estimated damages, ensuring its compatibility with the applicable regulations. Ensure that the document reflects a rigorous approach in legal terms, with a clear structure and aligned with international standards of information protection in corporate environments'.
Reviewing a contract and pointing out clauses that could represent legal risks for the contracting company.
Compare the key differences between employment regulations in Spain and France in relation to the dismissal of employees.
Summarise in 3 paragraphs the ruling of the Court of Justice of the EU on the right to be forgotten and its impact on digital privacy.
To perfect your use of legal prompting and get better results, consider the following tips:
If the response you get is not helpful, adjust the wording of the prompt by adding more detail or asking for a specific structure.
Saving examples of well-formulated prompts for similar cases will help you optimise future interactions with the AI.
Using specific terminology improves the accuracy of answers. Example: ‘Describe the principle of proportionality in criminal law with a case study.’
Make sure the question is not too general. A prompt such as ‘Explain data protection’ will give a vague answer, whereas ‘Explain how the GDPR regulates the retention of personal data in the EU’ will provide more useful information.
Legal prompting is an essential tool for lawyers and legal professionals looking to maximise the use of AI in their day-to-day work. From contract generation to regulatory analysis, knowing how to formulate the right prompts can make all the difference in the quality of the answers you get.
If you want to optimise your work with AI, mastering legal prompting will give you a competitive advantage in a constantly evolving industry. Now that you know the keys, are you ready to apply these strategies in your legal practice? If you want to learn advanced techniques from our experts and discover our legal generative AI assistant, Libra, contact us and ask for a free demo.