AI is beginning to change the business of law

In the spring of 2024, just two days after a complex cardiac surgery in the Midlands, a seventy-year-old man unexpectedly died. The case went to the coroner, and the deceased's family, seeking answers regarding the cause of the tragedy, hired Anthony Searle, a barrister specializing in medical negligence. When the coroner rejected the request to appoint an independent expert witness, Searle faced a wall of lacking resources in an underfunded legal system. He found the solution in technology that until recently was associated by lawyers solely with the risk of professional embarrassment: Generative AI.

The use of ChatGPT allowed Searle to prepare precise, technical questions regarding the course of the operation, filling the gap left by the absence of an expert witness. This is not an isolated case, but a signal of a deep shift in the legal sector. Although the profession has relied on tradition and slow, incremental evolution for centuries, artificial intelligence is beginning to penetrate its foundations – from preparing litigation strategies and automating compensation calculations to systemic reforms at the government level.

Precision instead of hallucinations: AI as a medical assistant

In the public consciousness, the legal use of AI is mainly associated with high-profile blunders, such as fabricated case law citations. However, professionals like Searle prove that the key lies in the proper direction of the tool. The barrister emphasizes that in his work, the law is usually an undisputed matter rooted in facts. The real challenge is medicine – understanding the nuances of a diagnosis, the course of a procedure, or the specifics of a particular treatment. This is where dedicated programs come to the rescue, such as PubMed integrated with the Claude model from Anthropic.

These tools allow for instantaneous searching of clinical studies and articles in medical journals, which drastically reduces the time needed to prepare for a hearing. Significantly, the medical experts cross-examined by Searle are often impressed by the substantive level of the questions he generates with the support of his "automated assistants." Searle maintains rigorous security standards throughout: he never enters client data into public AI models and meticulously verifies every piece of information and citation generated by the bot.

In addition to research, the technology is used to create dedicated solutions. Searle developed an application for calculating damages in medical negligence cases. This tool analyzes data from actuarial tables used by English courts, taking into account variables such as age or lost pension contributions, allowing for the generation of much more precise estimates of the victims' future financial losses.

Systemic revolution and government support

While individual lawyers experiment on their own, the British government sees AI as a cure for the chronic problems of the justice system, such as court backlogs and lack of funding. Proposed reforms involve implementing artificial intelligence for managing case lists, translations, and creating transcriptions. The scale of determination is shown by the fact that David Lammy, Secretary of State for Justice, delivered a key speech on judicial reform at a Microsoft AI event in London, which would have been unthinkable just a few years ago.

• Administrative automation: AI is intended to support court officials in managing calendars and negotiating fees.
• Drafting arguments: Consideration is being given to using technology to create so-called skeleton arguments, which are summaries of cases presented in court.
• Cost efficiency: Clients increasingly expect law firms to use AI to lower legal service costs.

The impact of technology is also visible in the structures of the market's largest players. The law firm Shoosmiths allocated one million pounds for bonuses for employees who reached the threshold of one million queries in the Microsoft Copilot system. Meanwhile, the American firm Ropes & Gray encourages junior lawyers to devote one-fifth of their billable hours to experimenting with technology. Some, like Clifford Chance, decided as early as late 2023 to reduce support staff by 10%, arguing, among other things, that some tasks were being taken over by automated systems.

Ethical barriers and professional risks

Despite the enthusiasm, the industry remains vigilant. Bruce MacEwen of the New York consulting firm Adam Smith, Esq., notes that although AI is the most frequently discussed topic in law firm boardrooms, evidence of a real, deep transformation of the multi-billion dollar sector is still limited. Leaders fear making binding decisions based on technology whose ultimate direction of development remains uncertain. The main deterrents are client data confidentiality and information protection, which limit the scope for safe use of public AI models.

Judges of the British High Court issued a stern ruling last year, condemning the use of false information generated by AI by barristers. In one case, as many as 18 fabricated case law citations were detected. "Artificial intelligence is a tool that carries both risks and opportunities. Abuses in this area have serious consequences for the administration of justice and public trust," reads the ruling from June last year.

Data from a LexisNexis study show a dynamic increase in technology adoption: by January 2026, as many as 50% of barristers declared using AI in their legal work, representing a twofold increase compared to 2024. However, only 2% of respondents claim that this technology is fully embedded in their operations and strategy. This suggests that most lawyers treat AI as ad hoc support rather than the foundation of their practice.

Ultimately, as Anthony Searle notes, law is a human domain, based on empathy and judgment that a machine cannot fully replace. AI can challenge expert opinions and optimize processes, but the final decision and responsibility for human fate must remain in the hands of the lawyer. The legal industry is moving toward a hybrid model where technology does not replace the advocate but becomes an essential condition for their effectiveness in an increasingly complex world of data.