Powered by
MOMENTUM
MEDIA
A bold prediction by Mustafa Suleyman that legal work could be fully automated within 18 months has sent ripples through the legal profession, with many legal leaders at the forefront of this technological shift insisting that the reality will be quite the opposite.
Predictions that AI could replace most legal work within the next two years are being met with growing scepticism from lawyers, many of whom insist that the profession’s core functions remain far beyond the reach of any machine.
The debate was sparked by comments from Microsoft AI chief Mustafa Suleyman, who warned that AI may soon reach human-level capability across many professional roles, with the legal profession among those most under scrutiny.
Speaking with the Financial Times, Suleyman cautioned that within just 12 to 18 months, AI could rival human performance across most professional roles – putting the everyday tasks of lawyers directly in the line of automation.
“I think that we’re going to have a human-level performance on most, if not all, professional tasks,” Suleyman said.
“So white-collar work, where you’re sitting down at a computer, either being a lawyer or an accountant or a project manager or a marketing person – most of those tasks will be fully automated by an AI within the next 12 to 18 months.”
While AI tools are increasingly finding a place in legal practice, many legal leaders argue that predictions of total automation dramatically underestimate the nuance, expertise, and judgement at the heart of the profession.
James Glissan, practice manager and solicitor at Glissan & Associates Lawyers, argued that such a perception of a fully automated legal work misunderstands the essence of the profession, which extends far beyond the mere mechanical application of rules.
“Predictions that legal work will be ‘fully automated’ within 18 months misunderstand what lawyers actually do. Artificial intelligence is already proving extremely useful for tasks such as document review, research, and drafting, and it will undoubtedly change how lawyers work,” Glissan said.
“But law is not simply the mechanical application of rules.
“It involves judgement, ethics, strategy, and navigating complex human disputes where facts are contested and the consequences are serious. Courts ultimately require accountability from a person, not an algorithm.”
Far from replacing lawyers, Glissan predicts AI may often lure people into trying to solve legal problems themselves, producing tangled and increasingly complex issues that eventually require professional expertise to resolve.
“In practice, what I am seeing is almost the opposite of the prediction. People increasingly rely on AI tools to try to solve legal problems themselves, only to create larger and more complicated issues that later require professional advice to untangle,” Glissan said.
Similar concerns have been voiced by Ryan Solomons, dispute resolution partner at RedeMont, who cautioned against the widely circulated view that lawyers’ work could be fully replaced by AI within just a few years.
Solomons maintains that while the growing use of large language models (LLMs) may streamline routine tasks, it will not make lawyers’ lives easier; instead, it will shift their focus to more complex work, making their roles more challenging rather than less demanding.
“In the next five years, legal work will get harder overall as the more mundane tasks get streamlined by LLMs, and the focus will be more on the more complex aspects of legal work,” Solomons said.
Although legal tech tools are frequently hailed as “intelligent”, Solomons pushed back, emphasising that the practice of law demands emotional skills and human judgement to tackle complex problems and deliver effective outcomes.
“LLMs also don’t do what lawyers do as LLMs do not think and are not intelligent. You need thought, empathy, and an ethical approach to solve legal problems and serve the proper administration of justice,” Solomons said.
Instead, Solomons explained that true intelligence requires learning, reasoning, and judgement shaped by real-world experience, qualities that LLMs, which he described as “autocomplete on steroids”, simply do not possess.
“Intelligence is the ability to learn, understand, and make judgements or have opinions that are based on reason and which can change based on your own experience with the real world,” Solomons said.
“LLMs don’t do that. LLMs are essentially autocomplete on steroids and statistical pattern machines that do not possess judgement of their own.”
Georgia Potgieter, senior data and AI governance specialist at QBE Insurance, has also pushed back against predictions of AI-driven automation in law, emphasising that the profession is a deeply human craft, where outcomes hinge on judgement, personalities, and the unpredictable complexities no machine can replicate.
“I’d argue the legal domain is unusually resistant to full automation,” Potgieter said.
“Primarily because the underlying data is deeply human: inconsistent, context-dependent, and shaped by imperfect decisions. Legal outcomes rarely follow clean binaries – rather, they depend on interpretation, incentives, personalities, politics and circumstance.”
While Potgieter acknowledged that AI could disrupt the traditional business model of law and chip away at the premium once attached to busywork, she stressed that the true heart of legal practice remains firmly beyond a machine’s reach.
“What AI will challenge, however, is the longstanding premium placed on busywork. Large volumes of document review, drafting, and procedural administration were never truly the essence of legal practice, yet they often commanded significant fees,” Potgieter said.
She added: “The good lawyer will not disappear. However the busywork surely will, and those who can offer nothing more.”
“As the era of charging a premium for relatively simple work may be drawing to a close, the profession, if anything, is likely to become more robust.”
