27/12/2025

Why AI Will Not Kill the Legal Profession – A Reply

This blog post was prompted by an opinion piece published on 16 December 2025 on Spectator Life under the title “AI will kill all the lawyers – a barrister’s warning”. Drawing on the views of an anonymous, highly experienced English barrister, the article argues that artificial intelligence will render the legal profession in its current form largely redundant in the foreseeable future. According to that assessment, this development is not speculative but already well underway.

As a criminal defence lawyer who deals daily with the practical, human realities of legal work, I do not believe this diagnosis should remain unanswered. Not because it is entirely misguided, but because its deliberate exaggeration obscures essential dimensions of what legal practice actually entails. This article therefore seeks to offer a reasoned counterpoint: not a defence of the status quo, but an attempt to place the debate on a more realistic and differentiated footing.

(Picture: Christian Vorhofen / Lukas Staffler)

Starting Point

The reason for this blog is an opinion piece published on 16 December 2025 on Spectator Life entitled ‘AI will kill all the lawyers – a barrister's warning’. The author draws on the assessment of an anonymous, experienced English barrister who believes that artificial intelligence will render the legal profession in its current form largely redundant in the foreseeable future. This development is not only inevitable, but already well advanced.
As a criminal defence lawyer who is confronted with the practical realities of the law on a daily basis, I cannot let this diagnosis go uncommented. Not because it is wrong in every respect, but because its exaggeration obscures central aspects of legal work. This blog article is therefore intended as a deliberate counterargument – not as a defence of the status quo, but as an attempt at a more realistic assessment.

Key points made by the barrister

The author essentially argues that modern AI systems are already capable of producing complex legal briefs, appeals and expert opinions faster, more cheaply and with at least the same level of quality as highly qualified lawyers. Since clients would no longer have any reason to pay many times more for human labour in the long term, economic logic would prevail. Legal activities would initially be automated step by step and eventually almost completely replaced; attempts by the industry to stop this through political or professional law would be doomed to failure. Consequently, the author warns young people against studying law at all.

An edgy description

Of course, these theses are deliberately exaggerated. The text uses dramatisation, simplification and polemics to attract attention and stimulate thought. And it succeeds in doing so. Especially within an industry that traditionally defines itself in terms of status, expertise and institutional security, the article comes across as disruptive. It forces self-reflection by radically questioning established practices and deliberately challenging the self-image of the legal profession.

At the same time, however, this exaggeration is also the weakness of the argument. The complexity of legal work is reduced to a single aspect in order to derive a general prognosis. This is precisely where the following counterargument comes in.

The fundamental error in reasoning: reduction to legal text work

The fundamental error in reasoning in this article lies in equating legal work with the production of legal texts. Pleadings, expert opinions and appeals are undoubtedly an essential part of a solicitor's work, but they are not its core. Legal work does not begin with writing, but with understanding.

Clients do not approach lawyers with clearly structured facts, but with fragmented narratives, emotions, fears and expectations. One of the central skills of legal work is to translate these everyday narratives into legally relevant facts. This translation is not a purely technical process, but a social process that requires trust, experience and empathy.

AI can analyse and generate texts, but it does not experience situations. It does not recognise nuances, power relations or unspoken motives. It processes input – it does not replace the conversation from which this input arises in the first place.

Law as a human conflict resolution mechanism

Law is not an abstract set of rules that functions independently of social relationships. It is an institutionalised system for resolving interpersonal conflicts. These conflicts are rarely purely rational. They are emotional, historically charged and socially embedded. That is why our legal system is still based on the principle that conflicts between people are also resolved by people.

The oft-quoted principle that ‘justice must not only be done, it must also be seen to be done’ (ECtHR, 17/01/1970, case of Delcourt v. Belgium, app. no. 2689/65, § 31) expresses this idea precisely. Justice is not only a result, but a process that must be comprehensible, transparent and conveyed in a humane manner. Courts are not data centres, but places of ritualised conflict resolution.

This is particularly evident in criminal law. Criminal proceedings are deliberately designed to be oral and direct. Guilt, responsibility and punishment are deeply human categories. The role of the defence lawyer is not limited to written submissions, but also includes presence, strategic litigation and the ability to react to dynamics in the courtroom. This dimension cannot be automated without changing the character of the proceedings themselves.

The big picture beyond the trial

The barrister's contribution also ignores the fact that a significant part of legal work takes place outside of court proceedings. Mediation, settlement negotiations, out-of-court dispute resolution and preventive advice are areas where it is not the ‘best’ legal solution that counts, but the most viable one.

Clients often do not want victory, but rather a conclusion. They want planning security, to avoid losing face or to maintain relationships. Such goals cannot be derived from legal texts, but can only be developed through dialogue. It is precisely the confidentiality, the strategic moment and the trust involved in such processes that argue against outsourcing to AI systems.

Agreement on the economic findings

Where we agree with the barrister is in his economic analysis. The classic logic of "billable hours" is coming under pressure. It is no longer acceptable to bill standardised audits exclusively on the basis of a large number of man-hours when digital systems can work faster, more consistently and more cost-effectively.

The traditional billable hour model in particular is reaching the limits of its acceptability. Even though it is formally transparent, many clients find it difficult to control. The actual total cost often remains unpredictable, and the cost risk lies solely with the client. Against this backdrop, flat fees are gaining in importance because they allow for cost control and clearly define expectations.

This development does not mean a loss of value for legal work, but rather a shift towards a more value-oriented approach. The decisive factor will be less the time spent than the concrete benefits provided.

Responsibility cannot be automated

The theory of completely replacing legal work also overlooks a key point: responsibility cannot be delegated. Even where AI is used, legal, ethical and liability responsibility remains with humans. Decisions must be accountable, liability must remain insurable, and trust requires personal responsibility.

An assistance model is therefore more realistic than a replacement model. AI is evolving from an aid to a highly developed assistant, a ‘co-pilot’ that analyses, structures and accelerates, but does not make decisions. This development is profoundly changing legal work, but it is not eliminating it.

Institutions, market and training

It is true that professional associations often take a defensive stance when it comes to technological upheavals. Monopolies and traditional business models are reluctant to be abandoned. At the same time, market logic cannot be permanently regulated. However, the consequence is not the end of the profession, but its differentiation.

This differentiation must also be reflected in education. In addition to legal dogmatics, skills that have long been considered ‘soft’ will gain in importance: strategic thinking, social competence, negotiating skills and personality development. Lawyers of the future will be less generalists and more specialised problem solvers with a clear profile.

Outlook

The essay ‘AI will kill all the lawyers’ is not a swan song for the legal profession, but rather a deliberately exaggerated wake-up call. It forces us to confront outdated assumptions and established business models. However, anyone who reads it as a definitive prediction fails to recognise the social and institutional dimension of law.

The future of the legal profession lies not in resistance to technology, but in its responsible integration. The legal profession will have to change – but it is precisely in this change that its continued relevance lies.

Author

This article is an opinion piece by Dr. Lukas Staffler. It reflects the author’s personal professional views. The contribution is intended to encourage discussion and critical reflection on current developments in law and technology. It does not constitute legal advice, an instruction, or a representation of third parties, and is not directed against any individual.