By Master Buckland (Rt Hon Sir Robert Buckland KBE KC)
Master Treasurer, Ladies and Gentlemen, distinguished guests, I am delighted to have been invited by Master Reader to deliver one of this year’s lectures. I do hope you are not having to suppress a collective groan about the topic I have chosen, because the two vowels A and I have become, in an alarmingly short span, one of the main preoccupations of commentators, writers and thinkers. On an almost hourly basis, we are bombarded with more news about the latest developments in machine learning, quantum computing and agentic artificial intelligence that it is tempting to let things wash over our heads and plod along as we have always done, hoping for the best.
I do not think that such an approach will do. As lawyers, judges and administrators of justice, trained to spring into action only when instructed or when asked to decide on issues in a case, we are in danger of taking a passive role, assuming that we will only be asked to address AI issues reactively, as and when administration of justice processes themselves deploy the emerging technology.
The reality is that the world around us is already full of AI-it is being used by governments, business and indeed all of us every day, very often without us even realising it. An increasing amount of material that we consider as lawyers is AI generated in whole or in part. Some of it will be misinformation, disinformation or deepfake. How well-equipped are we really to deal with these developments? And putting it bluntly, do we lawyers know how what we want, what we need and how we would like to work with AI?
These past few years have been, rather like in the case of the esteemed hobbit with whom I share a birthday, Bilbo Baggins, an unexpected journey for me, but in the time since I left Cabinet, I have found myself embarking on a path that has taken me across the Atlantic to Havard, across the United States meeting legal practitioners, academics at Stanford, tech bros and talented postgrads with whom I have had the pleasure of working, and now back again, this time to the Law School at LSE where I am in the very early days of my term as a Visiting Professor in Practice.
I have also been working with fellow lawyers at Payne Hicks Beach as we develop our own AI policies and ethical standards for its use in our work. At DAC Beachcroft, I have been seeing how the insurance industry, many of whom are clients of the firm, is already using AI to help assess applications and claims.
I have benefited from the wisdom of senior academics at the Harvard Kennedy School, and learnt to, if not love, at least respect, economists. At Harvard Law School, I enjoyed discussing and debating the ethics of AI with, amongst others, Professor Cass Sunstein (he of “nudge” theory and many things besides), whose speaking style is, if you haven’t seen it, most engaging and somewhat unusual. I am not going to copy him tonight by plunging directly into the audience, which given the layout of this excellent auditorium would be a rather eccentric and maybe life-threatening thing to do!
Richard Susskind rightly characterises the AI revolution as a transformational process in human evolution, not merely a bolt-on to our traditional work and life patterns. AI is changing the very notion of work itself, making tasks that once seemed impossible now within easy reach, or making human input into work largely or even entirely unnecessary. To those who say: what will become of us, my answer is simple: it is still up to humans and the choices they will make.
As AI research and development speeds up and the race between, the USA and China intensifies, we must ask ourselves whether we are making enough room for sensible choices or whether we are allowing too much risk to develop. Knowing human nature as I do, I am inclined towards the latter but firmly believe that we can manage this risk.
Today, I want to discuss the themes I have been pursuing in achieving a greater understanding of the impact of these changes on our justice systems and to propose a balanced path forward, rooted in the values of fairness, transparency, humility, and the dignity of being human. As this is such a fast-moving tableau, there is a risk of becoming a latter-day Dr Casaubon, whose academic research in George Eliot’s “Middlemarch” proved arid and obsolescent, as he was overtaken by intellectual developments elsewhere that were not revealed to him until it was far too late. This is why focusing on current technology and coming to confident conclusions based, for example, on the fact that many current Large Language Models regularly hallucinate, is unwise.
In my first Harvard paper, “AI, Judges and Judgement: Setting the Scene,” published in late 2023, I decided to focus, not on the technology itself, but about the very essence of justice, namely the nature of judgement itself. Precisely what is it that lies at the heart of justice and judgement, and what is it about that human element that makes us have trust and confidence in the system? If we are aligning machine systems with human values, then it is best that we are clear about what they are in the first place.
AI, in its most basic form, refers to machine systems capable of performing tasks that once required human intelligence, such as decision-making, pattern recognition, and predictive analytics. In the realm of justice, AI’s applications in justice currently range from facial recognition devices increasingly being deployed by the police, work on improving risk assessment tools used in sentencing and parole to e-discovery in litigation and the automation of administrative tasks within HMCTS.
More widely, the use of AI to determine minor consumer disputes, such as via PayPal for items bought online, is something readily accepted by millions of people. Certain state jurisdictions, such as China, have been progressively using machines to determine civil and consumer cases and have deployed machine learning to audit and vet human decisions, in a drive to create greater legal certainty and consistency.
AI offers the potential to enhance efficiency, reduce backlogs, and even correct certain human biases. But these technologies do not exist in a vacuum; they are created, trained, and deployed by people, and as such, inherit our values, assumptions, and failings.
When datasets are constructed with ethics in mind, we have less to fear, but all too often, they are not. In the case of China, data that does not advance the objectives of the Communist Party or socialist principles must not be used in the development of systems. This is deeply troubling for those of us who take issue with the Chinese system. The dangers of using fully open-sourced and often opaque datasets in the field of justice should be very clear. Not only is it a case of “rubbish in, rubbish out,” but also issues with data protection, privacy, and legal privilege.
We’re seeing technology that can process vast amounts of data, identify patterns, and even predict outcomes with remarkable speed. In the justice system, this could mean faster case resolution, reduced backlogs, and improved access to justice. AI has the potential to enhance fairness—at least in theory—by removing some of the inconsistencies and biases that human judges may bring to the bench.
But we must tread carefully. Judgement, in its truest sense, is not just about applying rules to facts. It’s about understanding context, exercising moral reasoning, and sometimes, showing empathy. In the paper, I distinguished between determinative judgement—where rules are clear—and reflective judgement, which requires deeper contemplation. AI may be able to handle the former, but it struggles profoundly with the latter. Invoking the wisdom of King Solomon, he didn’t just apply the law; he understood human emotion: “Solomon rightly decided that only the true mother would consent to her child being given away… This was an assessment of credibility based upon a shared understanding of basic human emotions.” Will we ever have full confidence in a machine reflecting emotion, as opposed to ourselves?
Bias is another critical concern. AI systems are only as good as the data they’re trained on. If that data reflects historical prejudices, the algorithm will replicate and even amplify those biases. We’ve seen this in systems like COMPAS in the United States, which have been accused of racial bias in sentencing recommendations. I identified four sources of bias in AI: algorithmic bias from flawed or prejudiced data, coding bias where laws are misinterpreted or oversimplified, misuse where systems are deployed for political or commercial ends, and human-AI interaction bias where judges either over-rely on or reject AI input without proper scrutiny.
Transparency is vital. Many of these systems are proprietary, meaning their inner workings are hidden from public view. This undermines due process and erodes public trust. We need explainable AI—systems that can justify their decisions in terms humans can understand. And then there’s the threat of deepfakes. These technologies can fabricate convincing audio and video evidence, casting doubt on the authenticity of real evidence. This leads to what scholars call the “liar’s dividend”—where even genuine evidence is dismissed as fake. We’ve already seen cases in the UK and US where deepfakes have disrupted legal proceedings.
I concluded by stating that the starting principle should be that AI should assist, not replace, human judges. We need robust frameworks to ensure transparency, accountability, and fairness. And we must prepare our legal systems to confront emerging threats like deepfakes head-on. The first paper set the scene, but in my second Harvard paper, which shares the title of this lecture, I sought to come up with some potential solutions that will be a framework for the use of AI in our justice systems.
A growing and real question for us now is: will people still want to use conventional court litigation systems if they can access private dispute resolution processes that are cheap and fast? Does increasing familiarity with AI mean that more people will readily consent to automated decision-making in justice? I think the answer is a resounding yes, but that the consequences for the existing system and our rule of law do not have to be a zero-sum game. Instead, state systems of justice can, at their heart, enshrine principles of fairness, human rights and independence of decision-making that will be the “kite mark” or gold standard of a justice system that has integrity. This will continue to be of particular importance when it comes to crime and punishment, family arrangements for children, and reputation-affecting disputes.
In England and Wales, reaching that gold standard in some respects is, frankly, proving difficult, however. The Crown Court backlog has almost surpassed 80,000 cases. Victims, witnesses, and defendants are waiting years for resolution. The justice system is under strain, and traditional methods are no longer sufficient. In my submission to the Leveson Review of Criminal Justice, I have strongly advocated the use of agentic AI at the earliest opportunity, with the automation of routine tasks, and assistance in evidence reviews that can ease the burdens for hard-pressed court clerks and judges. Agentic AI can help to create a culture of compliance with court orders, but efficiency must never come at the expense of fairness.
The digital age has transformed the nature of evidential material. Investigations now involve vast amounts of digital data—texts, emails, images, videos. Manually reviews are slow and costly. AI can help, but its deployment must be carefully governed. We must distinguish between assistive technologies, which support human decision-making, and automated adjudication, which can replace it. The former is already in use; the latter demands rigorous ethical scrutiny. This does not, however, mean that we should baulk at its introduction into the system, and at the earliest possible opportunity.
In 2023, guidance was issued urging judges in England and Wales to understand AI tools, avoid entering confidential data into public systems, and remain vigilant about bias and hallucinations. The Law Society’s 2024 guidance outlines risks—intellectual property, data protection, cybersecurity, and bias—and provides a checklist for ethical AI use. Tools like Harvey, Kira, and Spellbook are being developed for legal tasks. Yet adoption remains cautious due to concerns about reliability and hallucinations.
For those who look to governments for guidance, national and international regulation is fragmented at best, which means that moves to improve governance will have to come from the legal sector itself. Here in England and Wales, the Ministry of Justice published its Action Plan on AI just over two months ago, A Chief AI Officer and Justice AI Unit have been created, and there is a cross-departmental Steering Group.
Several initiatives are already happening, for example as part of the Digital Justice System initiative, an AI chatbot is being developed which supports users in resolving their child arrangement disputes, a policy area which has a vast and confusing landscape of information. Responses provide guidance on alternative routes to dispute resolution which could support a reduction in unnecessary court applications. The chatbot is trained on gov.uk content and uses AI persona scalable testing capabilities.
The MOJ states that it is also scanning potential for a single, secure identity for each person interacting with the justice system providing a joined-up view of individuals across services and enable more accurate, timely, and personalised support.
Further, the Action Plan contains some useful underlying principles, namely that AI in justice must work within the law, protect individual rights, and maintain public trust. This requires rigorous testing, clear accountability, and careful oversight, especially where decisions affect liberty, safety, or individual rights.
Secondly, AI should support, not substitute, human judgment. The independence of judges, prosecutors, and oversight bodies will be preserved, ensuring AI works within the law and reinforces public confidence.
Thirdly, AI tools should be designed around the needs of users, e.g. victims, offenders, staff, judges and citizens.
Finally, building common solutions that can be used across the system where possible, reducing cost and duplicated effort.
These statements are a welcome sign that maybe, just maybe, the mistakes of the past can be avoided. The deployment of older technology in parts of our justice system provides plenty of warning as to the dangers inherent in a “devolve and forget” approach.
The digital Single Justice Procedure, introduced for minor criminal cases some years ago, has improved efficiency but also raised concerns about fairness—especially for vulnerable defendants. This is a system that has retained human oversight until now, and even then, injustices have arisen with innocent people being wholly unaware of a conviction or written mitigation submissions just being ignored.
The Post Office Horizon scandal, although relating to what is now very old IT technology, revealed the dangers of opaque systems and lack of the ability to conduct scrutiny or to challenge the veracity of the flawed data that, compounded by misconduct, ended up in hundreds of wrongful convictions and monstrous injustice. These examples remind us of a simple truth: automation without oversight can lead to injustice.
What then, could be guiding Principles for the use of AI systems in the administration of Justice? In my paper, I set out six rules.
The first rule is Algorithmic Humility – AI must recognise its limitations and defer to human judgment in complex or sensitive cases. This rule stipulates that any AI system deployed in a judicial context must be programmed with an acute awareness of its own limitations. The system should be capable of recognising when it is operating at the edges of its training data or encountering scenarios that fall outside its realm of competence. For example, if an AI system processing a routine traffic offence detects language in the defendant’s statement suggesting mental health concerns or unusual circumstances, it should immediately flag the case for human review. This self-awareness is not just a safeguard against erroneous judgments, but a fundamental ethical requirement for any AI system entrusted with matters of justice. Incorporating this principle ensures that AI complements rather than compromises the integrity of judicial processes.
Context is crucial in judicial matters, as it allows for a deeper understanding of the circumstances surrounding each case, ensuring that justice is administered fairly and compassionately. In LLMs and generative AI, context is likewise extremely important, as it enables these systems to generate more accurate, relevant, and coherent responses by understanding the nuances and subtleties of the input they receive.
The second rule relates to the principle of ‘opt-in consent’, or “informed choice”. In the initial stages of AI integration, participation in automated judicial processes should be on a strictly voluntary basis. Defendants should be given a clear and informed choice between traditional human-led proceedings and AI-assisted adjudication, including the benefits and limitations of each. This opt-in approach serves several purposes.
Firstly, it respects individual autonomy, allowing those who are comfortable with AI systems to benefit from potentially faster processing times, whilst ensuring that those who prefer human adjudication are not forced into an automated system against their will. Secondly, it provides an opportunity for data collection and system refinement. By comparing outcomes between AI-processed and human-processed cases, we can continuously improve the accuracy and fairness of the AI systems.
However, the opt-in process must be carefully managed to avoid creating a two-tiered justice system. Clear information must be provided about the nature of the AI system, its decision-making process (often referred to as ‘algorithmic transparency’), and the rights of appeal. For example, a defendant opting for AI adjudication of a parking fine should be informed that while the process may be quicker, they retain the right to appeal the decision to a human judge if they are unsatisfied with the outcome. This transparency is important in maintaining public trust in the justice system.
The third rule in our framework is the principle of ‘contextual sensitivity’. AI systems must be capable of recognising and flagging cases where broader societal or systemic issues may be at play. For example, if an AI system processing traffic offences notices a statistically significant increase in speeding tickets issued at a particular location, it should not simply process these cases in isolation. Instead, it should flag this pattern for human investigation, as it may indicate issues with road design, signage, or faulty speed cameras. This ability to identify potential systemic issues is crucial in ensuring that AI systems do not inadvertently perpetuate or exacerbate existing inequalities or flaws in the justice system. Moreover, contextual sensitivity extends beyond just traffic offences.
For instance, in the health and social care sector, an AI system might detect a rise in certain medical conditions within a specific demographic. Rather than merely diagnosing and treating these cases individually, the system should flag this trend for further investigation to determine if there are underlying environmental, social, or economic factors contributing to the increase. This could lead to more effective public health interventions and policies.
Consider the case of an area where a disproportionate number of parking fines are being contested. An AI system processing these cases individually might miss the broader context – perhaps unclear signage or a recent change in parking regulations that hasn’t been well communicated. By flagging this pattern, the AI enables human authorities to investigate and address the root cause, potentially preventing unnecessary penalisation of residents and fostering a fairer application of the law.
This will be a distinct improvement on the current situation, which largely depends on the anecdotal experience of individual judges and is nowhere as near as comprehensive or systematic as an AI system of this nature. AI systems could also track the outcomes of contested fines to identify any biases in the adjudication process.
For example, if a particular demographic is more successful in contesting fines, this could indicate potential disparities in how fines are issued and contested. By flagging such trends, the AI system can help ensure that the enforcement of parking regulations is equitable and just. In essence, the AI’s ability to contextualize and analyse data from multiple angles ensures that if not only addresses immediate issues but also contributes to long-term improvements in policy and enforcement.
The fourth rule in our framework is the ‘principle of continuous human oversight’. While AI systems may be entrusted with certain decision-making processes, there must always be a clear chain of human responsibility and the possibility of human intervention. This rule mandates regular audits of AI decisions, random sampling of cases for human review, and clear processes for appealing AI decisions to human judges.
For example, in a system processing minor civil claims, a certain percentage of cases should be randomly selected for review by human judges, regardless of whether the parties involved have requested an appeal. This ongoing oversight serves to maintain the integrity of the system and provides a mechanism for identifying and correcting any systematic biases or errors that may emerge over time.
The fifth rule is the ‘principle of ethical transparency’. Any AI system deployed in a judicial context must be open to scrutiny, with its decision-making processes explainable in clear, non-technical language. This transparency is crucial not only for maintaining public trust but also for ensuring that defendants can effectively challenge decisions if necessary.
For instance, if an AI system recommends a particular sentence in a minor criminal case, it should be able to provide a clear explanation of the factors it considered and how they influenced its recommendation. This explanation should be comprehensible to the defendant, their legal representation, and the public. Additionally, ethical transparency entails regular audits and assessments of AI systems to ensure they remain fair, unbiased, and aligned with the principles of justice.
The sixth and final rule in our framework is the ‘principle of adaptive learning’. While AI systems must operate within strictly defined parameters, they should also have the capacity to learn and improve over time based on feedback from human oversight. However, this adaptive capability must be carefully managed to prevent the emergence of unintended biases or drift from established legal principles. Any significant changes to the AI’s decision-making processes should be subject to rigorous testing and approval by a panel of legal experts before implementation.
These six rules – algorithmic humility, opt-in consent, contextual sensitivity, continuous human oversight, ethical transparency, and adaptive learning – will, I believe, form a robust framework for assessing the suitability of AI integration in our justice system.
How, then, to deploy AI into the justice system itself? I believe that a Tiered Framework provides a clearest approach here. The first tier is Human-only adjudication, reserved for complex or precedent-setting cases, such as serious crimes and constitutional challenges.
AI-Assisted Human Adjudication supports moderate cases by aiding research and analysis, with humans making final decisions. AI can help detect bias or policy violations, but judges assess evidence and witness credibility.
The third tier, Human-Overseen AI Adjudication, allows AI a greater role in minor matters like traffic or small claims, always with human oversight for unexpected complexities. For example, complaints about unclear signage would be flagged for review.
Fully Automated Processing handles routine, uncontested cases, boosting efficiency. Even here, safeguards ensure appeal rights and regular audits maintain fairness. If AI detects sensitive issues, cases are escalated to human oversight.
Broad stakeholder scrutiny—by legal experts, ethicists, and technologists—helps evolve the system toward greater fairness. Regular review by joint boards ensures justice remains accountable and trustworthy.
Explainability of decisions
Integrating artificial intelligence (AI) into the justice system offers benefits but also presents challenges. Ensuring transparency and accountability in AI-driven legal decisions is crucial, though too much disclosure may enable individuals to manipulate outcomes.
If sentencing algorithms become fully transparent, defence representatives may exploit decision rules, potentially influencing trial outcomes—similar to strategies used in search engine optimisation. This raises questions of legislative intent and fairness. For example, if an AI model gives weight to community involvement and employment, legal counsel might advise clients to highlight these attributes, which could affect sentencing impartiality.
Such issues are also seen in financial services, where explainable AI credit scoring can lead to gaming the system. Ethical concerns about AI-assisted judicial processes highlight the need for robust oversight. Additionally, greater AI transparency in sentencing could reward coached displays of remorse or rehabilitation plans over genuine behaviour, possibly disadvantaging those with different cultural or personal expression styles.
I propose a system of “Calibrated Transparency,” which will offer varied levels of explanation: detailed breakdowns for judges and legal professionals, and general explanations for defendants and the public, all published in accordance with the vital principle of Open Justice.
Regular updates to AI parameters will help prevent gaming, just like changes to algorithms on social platforms. Monitoring behaviour patterns over time can distinguish genuine behaviour changes from last-minute attempts to influence AI. Human judges should provide reasoning, especially when diverging from AI recommendations, maintaining their role in the legal process. Training programs for legal professionals should stress ethical AI interaction rather than encouraging a gaming of the system.
This approach balances explainability and system robustness; transparency ensures fairness, but too much detail risks manipulation by the unscrupulous. Ongoing audits and adaptive measures are vital as new forms of gaming or imbalances emerge. The rights of defendants must be accounted for so legitimate defence strategies aren’t penalised.
Explainability techniques clarify AI decision-making by showing the importance of certain factors, helping in both scrutiny and understanding. Standardised explanation formats (eg in explaining how a criminal sentence was reached by reference to the relevant Guideline and then factors relevant to the particular case) would specify the hierarchy of different factors, quantitative influences, case comparisons, and outlier factors, supporting transparency, deterring gaming, and refining AI systems.
However, increased explainability might lead to even more sophisticated manipulation and an undue deference to AI by judges. To mitigate this, AI should remain advisory, with a human rationale required when diverging from AI recommendations.
Data compilation and use
In my paper, I dealt with the issue of compilation of training data for use by the system. Data compilers must work closely with developers to ensure context and nuance are reflected in AI algorithms. The judiciary’s involvement enhances credibility and legal compliance but may carry technical limitations and historical biases. Interdisciplinary commissions can counteract bias but will require significant resources. Public-private partnerships combine innovation and accountability but demand careful management of conflicting interests and transparency.
A quasi-open-source approach allows key stakeholders access to models and data for review without full public release, whilst balancing transparency and proprietary concerns. Third-party validation and continuous monitoring maintain quality and relevance, while robust testing ensures that models continue to adapt to evolving legal needs. An overreliance on synthetic data can degrade model quality, however, which highlights the need for ongoing human oversight and fresh, high-quality data.
I suggest a four-layer protocol for integrating AI into the justice system: Data Scrutiny, Model Transparency, Continuous Monitoring, and Societal Impact Assessment.
Data Scrutiny involves thoroughly analysing the training data used for AI models, considering not only demographics but also historical and societal context. Collaboration with social scientists and legal historians is recommended to avoid perpetuating outdated biases. Data Ethics Boards should oversee data collection, ensure diverse representation, and address gaps and quality concerns.
Model Transparency calls for making AI decision processes clear through “Legal AI Explainers” (LAIEs). These systems would translate AI reasoning into legally relevant explanations and help maintain judicial oversight. The development of LAIEs will require interdisciplinary expertise and must balance openness with security against manipulation.
Continuous Monitoring requires real-time bias detection using advanced neural networks that check AI outputs for emerging biases, adjusting data weights and ensuring decisions remain fair. All adjustments must be transparent and documented to support oversight and prevent new forms of bias emerging.
Societal Impact Assessment proposes establishing AI Justice Impact Units (AIJIUs) to regularly review how AI affects different communities. These units would combine statistical analysis and community engagement to identify disparities and recommend systemic changes beyond technical fixes.
This data framework, and indeed, all my suggestions, should be taken up by the AI Justice Unit. Ultimately, successful integration of AI in justice depends on constant vigilance, adaptation, and a commitment to fairness and equality.
Some Key Current issues
Let’s remind ourselves of some of the current challenges.
The dangers of synthetic media and deepfakes—realistic fake images, videos, or audio created by AI are clear. These pose risks to legal proceedings by making evidence authentication difficult, leading to potentially unfair trials or reliance on what should be inadmissible evidence. High-profile cases in the UK and US illustrate how deepfakes can undermine justice, either by introducing fabricated evidence or by allegations alone.
Existing evidentiary procedures are inadequate for addressing deepfake content. Proposed solutions include investing in self-authenticating technologies to detect deepfakes and assigning judges, not juries, the responsibility for evidence verification. However, these approaches face limitations due to technological costs and the rapid advancement of deepfake creation.
This is why I believe that law reform is needed to ensure that we criminalise the use of harmful synthetic content in courtrooms (updating the law of Perjury for example) and regulatory change to ensure that lawyers verify the authenticity of evidence before relying on it. I can also see a real utility in deploying blockchain technology to maintain a secure chain of custody for digital evidence.
Another challenge posed by AI is to future reliance on the court system itself. As AI becomes widespread in daily life, public trust and reliance grow, but this may lead to private dispute resolution via automated systems, making state courts less central for private cases. This shift raises concerns about the evolution of law, as public court cases contribute to legal development through accessible judgments. Will parallel dispute resolution procedures make state court processes in areas other than public criminal prosecutions increasingly marginalised?
Government use of automated decision-making, like benefits distribution or issuing documents, can affect accountability. Judicial Reviews rely on transparency (“Duty of Candour”) from government so decisions can be challenged; without explainable AI, important details may be hidden.
Enforcement of judgments is increasingly seen as suitable for automation. I agree and endorse the Civil Justice Council’s recent report which recommends a digital court portal for tracking finances and streamlining civil enforcement with judicial oversight to prevent abuse. Reliable databases and careful integration of AI are essential if we are to maintain fairness.
Conclusion
While AI is increasingly used in administrative and some civil legal tasks, justice systems should proactively establish ethical and professional standards for AI, with oversight by judiciary, legal professionals, and technologists. In England and Wales, whether through joint or multiple bodies, we must ensure that the “do no harm” principle is followed.
As a member of our Inn’s IT and AI Committee, I will do all I can to ensure that Inner Temple is at the forefront of thinking on this most exhilarating yet bewildering challenges. With a sense of humility and a willingness to engage in genuine collaboration, then AI can expand access to justice while upholding the rule of law- a true alignment that will retain that vital human element of justice.
Rt Hon Sir Robert Buckland KBE KC is a Governing Bencher of Inner Temple. He is Visiting Professor in Practice at LSE Law School, a barrister at Foundry Chambers, Senior Counsel at Payne Hicks Beach LLP and a member of the Policy Unit at DAC Beachcroft. He is the Third Church Estates Commissioner
