13 - 08 - 2026

When AI Meets the Corporate Lawyer

The machine can read faster than a lawyer, remember more than him and, if properly prompted, produce a neat little memo before the lawyer has finished his first cup of coffee.

But there is one thing Artificial Intelligence has not yet learnt to do: take responsibility for a bad legal decision. That, in essence, was the message from corporate legal expert Balanand Menon at a discussion on the changing face of the General Counsel.

AI, Menon said, is particularly good at gathering evidence. It can rummage through mountains of data, locate documents, sort information and present it with impressive speed. But deciding what all that information means for a company is another matter altogether.

The lawyer, not the machine, must decide where the danger lies.

Human judgement

Menon offered a homely analogy. AI, he suggested, is rather like a very efficient chef. It can collect the vegetables, wash them, chop them and arrange them neatly on the kitchen table. What it cannot yet do is decide whether the customer wants a curry, a salad or something fit only for the dustbin. That final judgment belongs to the human being.

This distinction may sound obvious. In the corporate world, however, it is becoming increasingly important. General Counsels are being pressed to embrace AI, but they are also being asked the most irritating question in corporate life: “What is the return on investment?”

Saving a lawyer 45 minutes by producing a summary is useful. But it is not necessarily a revolution. The real test is whether those 45 minutes translate into something the company can recognise on its balance sheet — better decisions, wider legal coverage, fewer risks or lower costs.

And therein lies the rub. AI is moving from being a clever personal assistant to becoming part of the machinery of the legal department. It can help review marketing material, assist with compliance, sift procurement documents and even operate through specialised agents.

The promise is enormous. A legal department that once had to choose which problems it had the manpower to attend to may soon be able to keep an eye on many more corners of the organisation.

But there is a catch. Machines are wonderfully confident creatures. They can produce an answer even when they ought to say, “I don’t know.”

A young lawyer who treats an AI-generated document as gospel may therefore save an hour and cost his General Counsel a sleepless night.

The problem is not that AI makes mistakes. Humans have been making mistakes since Adam blamed Eve. The problem is that AI can make mistakes at extraordinary speed and present them in impeccable prose.

That is why, the panel argued, the lawyer of tomorrow cannot remain merely a specialist in statutes and case law. He will have to understand technology, data, workflows and the business itself.

The emerging model is that of the “T-shaped” lawyer — deep knowledge in one legal discipline, with a broad understanding of everything else that makes a modern corporation tick.

There is another uncomfortable truth. AI is changing so quickly that yesterday’s training may become tomorrow’s archaeological artefact. Lawyers will need to keep learning, experimenting and questioning the machines they increasingly work alongside.

The future, therefore, is unlikely to be a courtroom drama in which a robot replaces the lawyer and everyone goes home.

It is more likely to be a less dramatic, but far more consequential partnership: the machine searches, sorts and summarises; the lawyer asks the inconvenient questions, understands the business and decides how much risk the company can afford to take.

The machine may know more. The lawyer must still know better.