Authored By: Favour Chukwudum
Manchester Metropolitan University
Introduction
Family Law dispute deals with emotionally complicated situations, such as divorce, child arrangements, domestic abuse, and financial settlements. In England and Wales, getting timely legal counsel is still a constant difficulty, despite the importance of these issues. The reduction of legal Aid, Sentencing and Punishment of Offenders Act 2012 (LASPO), increases court delays, and an increase in the number of litigants in person.[1] Because of this, a lot of people look for legal information on their own frequently using unreliable and low-quality web sources.
Large Language Models (LLMs), one of the most recent developments in artificial intelligence, offer fresh approaches to these problems. Before hiring a lawyer, LLM-based triage system can help users by outlining legal processes. Pointing out pertinent legal concerns, and pointing people in the direction of the right resources. Even though these technologies have a great deal of potential to increase access to bring justice, they also bring up significant issues with professional ethics, accuracy, accountability, and secrecy. This article argues that LLM-based triage systems can significantly improve access to justice in family law provided they operate as decision-support tools under meaningful human supervision and within an effective regulatory framework. It first examines the contribution of LLM-based triage to improving access to justice before considering the regulatory and professional safeguards necessary to ensure their responsible use.
LLM-Based Triage Systems: Opportunities for Access to justice
Large Language Model (LLM)-based triage systems represent one of the most significant developments in the use of artificial intelligence within legal services. Rather than replacing lawyers, these systems are designed to provide an initial assessment of a user’s legal issue by analysing information provided in natural language, identifying relevant legal categories, explaining possible options, and directing individuals towards appropriate sources of assistance. In this way, by removing the initial obstacles that many people encounter while trying to comprehend their legal rights, LLM triage tools may enhance access to justice.[2] AI-assisted triage is especially useful in family law since disputes often involve people who need immediate assistance but may not be well-versed in the law. Early knowledge of the legal options is frequently necessary for issues including divorce, child arrangements, financial remedies, and protection against domestic abuse.
However, people may find it challenging to find the right legal assistance due to the emotional and complicated nature of family conflicts. LLM-based solutions can help users discover pertinent concerns, give clear explanations of legal procedures, and direct them to experts when needed.[3] A key benefit of LLM triage is is the potential to provide faster access to legal information. When demand outpaces available resources, traditional legal service can result in delays before people can receive counsel. Basic legal questions can be answered instantly by AI systems, giving people a greater understanding of their situation before requesting more help. This is in line with HM Courts and Tribunals Service’s (HMCTS) larger push for digital justice services, which aims to update legal procedures using technologically enabled alternatives.[4]
Additionally, by helping people gain a basic understanding of the law before seeking professional assistance, LLM-based triage may help lower legal expenses. These technologies may enable lawyers and consulting firms to concentrate their limited efforts on more complicated matters by managing routine information requests and guiding users properly. This is especially pertinent in light of the strains on legal aid, where rising demand and scarce resources have made it extremely difficult to obtain counsel.[5] However, these systems should operate as supportive tools rather than substitutes for professional legal judgement. Solicitors utilizing artificial intelligence are nonetheless accountable for maintaining accuracy, confidentiality, and adherence to professional standards, according to the Solicitors Regulation Authority (SRA).[6] In similar vein, research by Izzidien, Sargeant, and Steffek shows that when technology is developed with user requirements in mind, it can improve access to justice. However, it must preserve the necessary protection to prevent disinformation and exclusion.[7]
Therefore, by increasing accessibility, lowering costs, and relieving pressure on current legal aid system, LLM-based triage systems have a significant chance to increase early legal assistance, especially in family law.
Risks, Regulation and Professional Accountability
Despite the potential advantages of LLM-based legal technologies, there are serious dangers associated with their usage in legal services with regard to professional accountability, accuracy, fairness, and secrecy. These issues are especially crucial in fields like family law, where giving the wrong counsel could have detrimental effects on those who are already vulnerable. Adoption of artificial intelligence must therefore be supported by suitable regulation and ongoing human control, even though technology may enhance access to justice. The generation of false information, sometimes known as “hallucination,” is one of the biggest concerns connected with LLM systems.
LLMs may give convincing but inaccurate legal explanations, references, or procedural guidance since they generate responses based on patterns in training data rather than true legal reasoning. Such mistakes are especially troublesome in a legal setting since users could base their decision about their rights and responsibilities on false information.[8] The risk is heightened for individuals who lack knowledge and may be unable to identify whether AI-generated advice is reliable. AI systems may also reproduce or reinforce existing biases within legal data. LLMs may reflect historical inequalities or inadequate representations of particular communities because they learn from vast datasets generated from human-produced data. This raise worries that vulnerable people may inadvertently be disadvantaged by automated legal assistance, especially when choices involve delicate topics like immigration, family disputes, or public service access.[9]
Confidentiality and data security constitutes another significant issue. The use of AI technologies raises concerns about how sensitive information is stored, processed, and shared, and legal professionals have strict obligations to protect client information. The SRA has emphasized that lawyers are still in charge of making sure artificial intelligence is applied in accordance with professional duties, such as requirements for competence, secrecy, and supervision.[10] Concerns about the unlicensed practice of law are also raised by the extensive use of AI legal tools. While AI systems might offer generic legal information, there is a difference between helping users comprehend legal matters and offering tailored legal advice that calls for expert judgement. This divide may be blurred by the deployment of AI without sufficient protections, potentially exposing people to untrustworthy assistance while compromising professional standards.[11]
The growing awareness of these concerns is reflected in regulatory initiatives. A risk-based framework for regulating AI systems is established under the European Union Artificial Intelligence Act, which includes criteria for accountability, transparency, and human oversight.[12] Professional regulators like the SRA have emphasized that technology must work within current legal and ethical requirements, even though the UK has taken a different regulatory strategy.[13] Concerns about the responsible application od AI in legal practice are further highlighted by recent remarks by Ayinde and Munir. They contend that even while AI technologies could boost productivity, legal practitioners still have to take responsibility for choices and results involving AI-assisted procedures.[14] This supports the idea that rather than taking the role of expert knowledge, artificial intelligence should improve legal services.
Therefore, the future role of AI within legal practice should be viewed as supportive rather than substitutive. Although LLM systems can help lawyers with regular work, efficiency and accessibility, human lawyers are still necessary for applying legal judgement, comprehending unique situation, and upholding ethical duty. Instead of using AI to replace lawyers, a cooperative relationship where technology enhances professional legal services while upholding responsibility and public trust is the proper course of action.
Conclusion
AI has the potential to drastically alter access to justice by increasing the effectiveness, affordability and accessibility of legal services. In fields like family law, where people frequently need prompt assistance, LLM-based systems may offer beneficial support through early legal information, case triage, and help with mundane duties. These advantages must, however, be weighed against serious hazards, such as biased results, confidentiality issues, and ambiguity around professional accountability.
This article has stated that AI should not be considered as a replacement for attorneys but as a technology that might enhance legal practice when properly regulated. Maintaining human supervision, guaranteeing adherence to professional obligations, and upholding public trust in legal services are all necessary for AI to be effective in the legal system.[15] Further adoption should therefore focus on regulated deployment, mandatory human review of AI-assisted legal work, and continued guidance from courts, regulators, and professional bodies.[16] By adopting this strategy, technology may uphold the core principles of justice, accountability, and fairness rather than erode them.
BIBLOGRAPHY
Primary Source
UK Statutes
Legal Aid, Sentencing and Punishment of Offenders Act 2012
EU Legislation
Regulation (EU) 2024/1689 Artificial Intelligence Act [2024] OJ L1689
Secondary Sources
Books
Susskind R, Online Courts and the Future of Justice (Oxford University Press 2019)
Susskind R, Tomorrow’s Lawyers: An Introduction to Your Future (3rd edn, Oxford University Press 2023)
Journal Articles
Ayinde and Munir, ‘Artificial Intelligence in Legal Practice: Opportunities, Risks and Professional Responsibility’ (2024)
Izzidien D, Sargeant N and Steffek F, ‘Artificial Intelligence and Access to Justice: Opportunities and Challenges’ (2023) International Journal of Law and Information Technology
Official Publications & Reports
HM Courts and Tribunals Service, HMCTS Reform Programme: Progress Update (2024)
Solicitors Regulation Authority, Artificial Intelligence and the Legal Sector (SRA 2024)
[1] Legal Aid, Sentencing and Punishment of Offenders Act 2012; Richard Susskind, Tomorrow’s Lawyers: An Introduction to Your Future (3rd edn, Oxford University Press 2023) chs 14–16.
[2] Susskind (n 1)120–124.
[3] Daniel Izzidien, Nicola Sargeant and Felix Steffek, ‘Artificial Intelligence and Access to Justice: Opportunities and Challenges’ (2023) International Journal of Law and Information Technology (discussing the role of AI tools in improving legal accessibility and user support).
[4] HM Courts and Tribunals Service, HMCTS Reform Programme: Progress Update (2024).
[5] Richard Susskind, Online Courts and the Future of Justice (Oxford University Press 2019) 87–91.
[6] Solicitors Regulation Authority, Artificial Intelligence and the Legal Sector (SRA 2024).
[7] Izzidien, Sargeant and Steffek (n 2).
[8] Susskind (n 1) 126-129
[9] European Union, Regulation (EU) 2024/1689 Artificial Intelligence Act [2024] OJ L1689, arts 10, 14.
[10]Solicitors Regulation Authority (n 6).
[11] (n 3).
[12] Regulation (EU) 2024/1689 Artificial Intelligence Act [2024] OJ L1689, arts 13–14.
[13] Solicitors Regulation Authority (n 6).
[14] Ayinde and Munir, ‘Artificial Intelligence in Legal Practice: Opportunities, Risks and Professional Responsibility’ (2024).
[15] Solicitors Regulation Authority (n 6).
[16] Regulation EU (n 12).
