Sheffield Digital Justice: Developing a Client Interviewing Tool
Kathleen Calamohoy, Olivia Hawkes, Violet Keen, Thanyha Kirisanker, Paige Seymour,, Ali Al Shuwaili
The importance of client interviewing
Client interviewing is the cornerstone of effective legal practice, serving as the essential foundation for gathering facts, identifying specific needs, and also establishing a trusted professional relationship with your client. Interviewing allows the client to be listened to and give their side of the story, which is essential, for example, if they have been implicated in a crime.
By creating a safe environment where a client feels truly heard and understood, lawyers can build the rapport necessary to encourage an honest, free flow of information regarding the client’s problems, concerns and feelings. The first initial meeting is critical for eliminating ambiguity and clarifying the client’s ultimate goals. Often what a client seeks (for example, revenge) must be carefully aligned with the realistic outcomes that the legal system can provide. Using an effective technique, such as starting with broad, open questions to capture the narrative before narrowing down to specific details, allows legal professionals to identify potential conflicts of interest and explore viable solutions early in the process. Ultimately, a successful client interview helps transform an emotional narrative into a clear, actionable legal strategy.
Skills required for client interviewing
The skills required for client interviewing include, among others, creating a safe space, active listening, building strong rapport and empathy, and effective, open-ended and structured questioning.
- Establishing a safe space for the client to share and disclose personal and sensitive information is crucial not only during but most especially prior to the start of the actual interview because without it, the possibility that clients would hesitate to give the necessary information is high.
- Active listening entails focusing on the client and making them feel that all their concerns, feelings and vulnerabilities are being taken into account. Legal professionals must be judgement-free and a good listener, as it makes the client feel more comfortable to open up. This helps to build a good relationship between the client and interviewer, which can have a positive effect on the client if they feel that they are in a safe space and can trust them with personal information.
- Building strong rapport and empathy during client interviewing pertains to the matter of trust. In order for the client to be comfortable enough to share the truth about their situation and freely express whatever concerns they may have, it is important for legal professionals to have and continuously develop this skill.
- Developing an effective, open-ended and structured way of questioning plays an important role in client interviewing because it allows legal professionals to gather all the details and information about the matter at issue without any limits, organize how the interview should go, and identify what is relevant or necessary to solve or address the client’s concern.
Developing client interviewing tools
During this project we each developed our own AI chatbots. This involved using the Wonda platform, which allowed us to input prompts for our chatbot, to act as a client requiring advice on a legal matter. In doing so we each developed a legal scenario, adapting our chatbot’s appearance, emotionality and voice to suit their legal issue and situation. Once they were created we made a spreadsheet with relevant evaluative points on important elements such as: speed and efficiency, detail, accuracy, understanding, reliability, feedback and visual appearance. We used this to record our findings when testing each other's chatbots, allowing us to make the required adaptations for the best output of each chatbot to reflect a real client interview scenario.
Examples of client interviewing tools
For the project, I developed Zeus, a 360° VR digital concierge within the Wonda VR platform to assist international students with campus navigation. My work focused on technical interaction logic, where I customized AI prompts that allow Zeus to listen for specific keywords and respond clearly and accurately. I focused on the ‘Digital Justice’ aspect of physical and administrative accessibility. The results were successful we proved that AI chatbots can effectively simulate the unpredictability of real-life clients, and my implementation of spatial markers showed how technology can reduce the stress of wayfinding for students in a new environment, in which I created a tool that makes complex university systems easier to navigate for everyone.’ Ali Al Shuwaili |
‘For my project, I made a Wonda AI chatbot that represented a young boy who had been implicated in a crime. I added information around what happened, and where the client was to create a story. My chatbox was set in a police station, where the boy was being questioned in regards to the murder that took place. Essentially you are acting as his lawyer to determine his innocence. We filled out an evaluation form of each other’s chatbot to help overcome any issues that we may have had and gain feedback in regards to our chatbot. I found that this chatbot worked really well and there were endless possibilities of how you could create it. As long as you gave enough information in your chatbot, then it would answer the questions like a real client interview. I think that the setting acted like a real police cell, which made it seem even more realistic.’ Paige Seymour |
‘For the project, I designed my chatbot not as a client but as a lawyer. I tried to create prompts which from the perspective of a lawyer could be helpful in client interviewing skills like what to do or how to communicate with the client before, during and after the client interview.’ Kathleen Calamohoy |
‘For this project, I developed the persona of ‘Elena,’ a 34-yr old survivor of domestic abuse. I tried to make it realistic by creating complex psychological layers into the AI’s system. This included, being really withdrawn and defensive, where the chatbot only reveals information such as being strangled or coercive control, if the user is demonstrating a specific empathetic behaviours and guarantees confidentiality, which will help lawyers learn how to communicate well with people going through similar situations. The chatbot is also developed to react to ‘bad’ legal habits, such as victim-blaming or using excessive legal jargon, by withdrawing or becoming defensive. This will allow lawyers to refine their interviewing techniques, ensuring they can identify life-threatening risks while maintaining the psychological safety of a vulnerable client. The results of the project was very positive as it demonstrates that empathy is a legal skill and not just a personality trait.’ Thanyha Kirisanker |
The opportunities and challenges involved in using AI to train law students and lawyers in client interviewing
The use of AI to train law students in client interviewing skills can provide several opportunities, but there are also notable challenges.
The use of AI offers an adaptable training tool that can be programmed to match the interviewer's skills. It allows users to develop the questions which you ask clients and build a relationship with them. As you can change the settings and use your own prompts, then this would be useful to use in any situation and help practice interviewing in any setting.
Unlike traditional client-interview training, which provides a uniform teaching method to all students, AI chatbot systems can be designed to adjust both the complexity of legal issues as well as the behaviour of a client. It can be programmed to simulate a wide range of legal scenarios and client personas, from cooperative individuals to aggressive or emotional clients, offering a tailored experience that traditional training cannot easily replicate. For example, a beginner may interact with a cooperative client who provides clear information about their legal issue. Whereas an individual who is more advanced could interact with a client who is more emotional and aggressive, as they have a harder legal issue. Such scenarios may be difficult to simulate in other training avenues. As a result, AI can provide a tailored learning experience, which may be difficult to achieve in the traditional learning setting due to time and resource constraints. Additionally, AI can provide learners with targeted feedback on their interviewing skills. Raising issues such as tone, missed follow-ups, and weak leading questions.
However, there are challenges that limit the effectiveness of an AI training tool. AI chatbots are often limited by their training data and specific scripts, they may not be able to handle complex, multi-sector legal issues that fall outside their programmed parameters.
A key concern is that the AI-generated chatbot would be too structured and predictable, not being able to realistically simulate a real client interview. As AI systems rely on pre-programmed patterns, they may respond in ways that feel robotic. They do not have real human feelings, so will follow the prompts given. This is unlike real clients who are forgetful, emotional and can often tell contradictory information. As such, students may adapt their interviewing techniques to the chatbot, which would result in questioning that is tailored to fit a system, rather than developing skills that would be appropriate within a real client interview. Since you know what prompts you have made, it is easier to ask the ‘right’ questions to get the answer you want, this will obviously be more difficult in a real setting. This could limit a student's ability to handle the ambiguity or unpredictability that is a part of a real client interaction.
An additional limitation is that our chatbots operate according to a script, dictating their personal characteristics and specific legal problems. While a chatbot may effectively simulate one type of legal problem, it would not be able to replicate different legal problems from other sectors of law. Consequently, in order to expose students to a multitude of different legal problems, additional chatbots would need to be created.
Learning from this project
‘The creation of my chatbot taught me the importance of being specific with prompts when using AI tools, to ensure the output reflects the input for each individual situation. As well as this the project showed me how AI tools can be used to practice interviewing, the use of the chat bots feedback can be applied to real life situations, which is vital in the legal profession to know how to deal with certain client situations. For example, the emotional aspect of each chatbot demonstrates how each client is different, therefore they each require different responses depending on their emotional state and legal situation. Overall, this project taught me the usefulness of AI tools in the legal profession, as they can be used to practice and perfect interviewing skills such as empathy and legal advice for different situations.’ Violet Keen |
‘Throughout this project, I developed a clearer understanding of the client-interviewing process and the relevant skills needed to interact with clients. As a law student, I was initially hesitant to volunteer for pro bono schemes because I lacked confidence in my communication skills. However, this experience and building a simulation have taught me questioning techniques and ways to build rapport with clients. Increasing my confidence to pursue opportunities with a real client. Additionally, through building the chatbot, I developed a better understanding of how digital tools can be used to simulate professional interactions. Learning how to structure prompts in a way that guides an AI’s system behaviour. This process has strengthened my problem-solving skills. A key challenge I encountered with my chatbot was that it initially defaulted to AI assistant responses, rather than remaining in character as a client. This required extensive testing and refinement to ensure my chatbot remained in character at all times. Overall, this project improved my client interviewing skills and understanding of how communication and AI simulations can support effective legal training.’ Olivia Hawkes |
‘Through this project, I learned that conducting a successful interview with a client going through domestic abuse requires more than just a legal checklist; it requires a sensitive, supportive approach that acknowledges the client’s emotional state. I discovered that practicing with an AI chatbot gives lawyers a safe way to improve their empathy and listening skills, especially when a client shares a story in a fragmented way. For example, if the AI becomes withdrawn or silent when asked a rude or blunt question, the lawyers will learn to pivot from this approach to using a softer, open-ended validation. This is incredibly beneficial as it will allow them to build up on these skills and learn from mistakes in this environment, ensuring they can uncover complex issues like coercive control and gather accurate evidence without re-traumatising vulnerable clients. Thanyha Kirisanker |
‘I learned from this project that for an effective and successful client interviewing tool, it should as much as possible simulate the real-world legal environment including being sensitive to the client’s concerns, accurate in the information needed to obtain from them, detailed and clear in the methods of communication, and empathetic to the client’s plight. AI can be designed to facilitate this and to assist lawyers and legal professionals to do client interviewing in an efficient way.’ Kathleen Calamohoy |
‘I learned that for a VR-based Digital Justice tool to be truly effective, the AI must bridge the gap between verbal dialogue and physical space. Through developing the Zeus in Wonda VR, I discovered that international students benefit most when AI conversations trigger immediate, visual cues. Specifically, I learned how to engineer interaction logic so that when a student asks for directions and clicks on text a 3D spatial marker is automatically revealed. This reduces the cognitive load of navigating a complex campus in a second language. This project taught me that legal technology isn’t just about drafting documents, it’s about using technology and spatial anchors to ensure that information is accessible and easy to find for everyone.’ Ali Al Shuwaili |
‘From this project I learned that AI can be used in really helpful ways for people aspiring to be lawyers. It can be used to help gain confidence and experience in client interviews, which is an important practice for those wanting to go into law. It can be used in multiple settings and scenarios, which is beneficial if you are unsure what area of law you want to go into.’ Paige Seymour |
Conclusion
The results of this project were positive, the chatbots are a great way at practicing client interviewing, especially because they allow users to practice giving advice on different legal areas. As well as this, a number of the chatbots were prompted to act emotionally, which allowed users to practice real life client interview situations where a client may be stressed or angry, which is a crucial skill when interviewing real clients.
The testing phrase of the project enabled the chatbot creators’ to practice their own interviewing technique, giving feedback which can be applied and used in later real life interviews. The use of the evaluation spreadsheet when testing the chatbots ensured they were at the required standard to be used in a client/ interviewer situation. Using this as a method of peer review significantly improved the chatbot creators’ understanding of testing and evaluating client interviewing methods, as well as enabling them to practice required prompts for each AI chatbot. allowed
Overall, this project effectively demonstrated how the use of AI chatbots can reflect real-life client interview situations, which are highly significant in the legal profession.