Precedent meets pace.

The law is the only profession that builds its future entirely out of its past. Every argument leans on precedent, every contract on a clause that worked before, every piece of advice on the accumulated judgement of those who came earlier. It is a discipline engineered to move carefully — and for good reason. Carelessness in law is not inefficiency; it is malpractice.

So it is no surprise that legal practices have approached artificial intelligence with their arms folded. The scepticism is warranted: a tool that mishandles privileged material or invents a citation is worse than no tool at all. But while the profession debated, AI quietly stopped being theoretical. It is now inside the working day of forward-looking firms — comparing, reviewing, drafting, and surfacing insight at a pace no associate team can match.

The real question has changed. It is no longer whether the technology works, but whether a firm will adopt it on its own terms, or have the new pace dictated by clients who have already stopped waiting.

What AI is doing in law practices today

The most useful way to understand the present moment is to look not at what AI might eventually do, but at the concrete, unglamorous work it is already handling well.

Contract analysis and review. Where a lawyer once read a fifty-page commercial contract clause by clause, AI now reads it in seconds — flagging unusual terms, missing protections, inconsistent definitions, and deviations from a firm's preferred positions. The lawyer's role does not disappear; it elevates. Instead of spending the first two hours finding the issues, they spend those hours deciding what the issues mean.

Due diligence at scale. This is where the change is most visible. Consider a transaction involving ten commercial contracts that must be reviewed together — different counterparties, different governing terms, different renewal and termination mechanics. Traditionally this means a team, a deadline, and a great deal of manual cross-referencing. AI can now ingest all ten simultaneously and produce a structured comparative table: party, term, value, key obligations, termination rights, anomalies. What took days of associate time becomes a reviewable draft in minutes — one that a senior lawyer can interrogate, correct, and trust as a starting point rather than an endpoint.

The knowledge bench. Every established firm sits on decades of accumulated expertise — precedent files, prior advice, internal memoranda, negotiated positions. Most of it is effectively inaccessible, locked in folders no one has time to search. AI turns that dormant archive into a living resource: a lawyer can ask a question in plain language and receive an answer grounded in the firm's own historical work, with the source documents cited. The institutional memory of the practice becomes queryable.

Drafting and first-pass production. From engagement letters to standard clauses to client correspondence, AI handles the first draft — not to replace the lawyer's voice, but to remove the blank page. The work that follows is editing and judgement, which is where the value has always been.

What matters more than the capability

It is tempting to stop at the list of features. But the legal profession has specific, non-negotiable concerns, and any honest discussion of AI in law has to meet them directly.

Confidentiality and professional secrecy — sigilo profissional — are not optional. A tool that processes privileged client material in an uncontrolled environment is not an efficiency gain; it is a liability. The firms adopting AI responsibly are the ones insisting on data residency, controlled access, clear audit trails, and infrastructure that keeps client information within defined boundaries rather than feeding it into open systems. The technology is only as trustworthy as the architecture around it.

Equally, AI in law is not autonomous and should not be sold as such. It produces drafts, comparisons, and candidate answers. A qualified lawyer remains accountable for every output that reaches a client. The firms getting this right treat AI as an exceptionally capable junior — fast, tireless, and well-read, but always supervised.

What is coming down the road

The trajectory is reasonably clear, even if the timing is not.

In the near term, expect AI to move from assisting on discrete tasks to handling connected workflows — taking a matter from intake through document gathering, analysis, and a structured first draft of advice, with the lawyer stepping in at the decision points rather than the mechanical ones. Expect tools that live inside the firm's own systems rather than demanding that lawyers work in yet another separate application.

Further out, the competitive landscape itself will change. As AI absorbs the routine, the differentiator between firms will increasingly be the quality of judgement layered on top of it — and the speed with which a firm can turn its own accumulated knowledge into client value. Clients, particularly sophisticated corporate ones, will come to expect the speed AI enables. A due diligence summary that once justified a week of billable time will be expected within the day.

Firms that cannot meet that expectation will not lose on quality; they will lose on relevance.

Why a closed industry knows it must ride the wave

Legal practice is, by design, a conservative profession. That conservatism is a feature — it protects clients, upholds standards, and resists fashions that would compromise rigour. It is entirely reasonable for a firm to refuse to adopt a technology it does not understand or cannot control.

But there is a difference between caution and paralysis. The pressures pushing law toward AI are not coming from technology vendors. They are coming from clients who increasingly use these tools themselves and wonder why their lawyers do not; from younger lawyers who will not accept doing by hand what a machine does in seconds; and from competitors who have quietly decided that being first matters.

The firms that thrive will not be the ones that adopt AI fastest and least carefully. They will be the ones that adopt it deliberately — choosing tools that respect confidentiality, that integrate with how the firm already works, that keep the lawyer firmly in command, and that turn the firm's own expertise into an advantage rather than handing it to a generic public model.

A profession built on precedent has no precedent for this.

But that has never stopped good lawyers from exercising judgement under uncertainty — it is, after all, the work. The choice left is not whether the pace of legal work will change. It is who sets it.