Legal: full guide · AI for Executives

How each claim is supported:

EstablishedPublished resultSurveyVendor claimIllustration

Tap any marked phrase for its source.

Section 1 of 9

How this industry makes money

In one sentence

A law firm makes money by selling its lawyers' time, leaning on juniors to do much of the work, and collecting what it bills. AI changes all three at once.

Most firms sell time. Revenue is hours worked, times the rate, times the share of bills the client actually pays. Profit comes from leverage: juniors do much of the work at lower cost, and partners earn the margin on it.

Five pools of money decide a firm's year.

Pool	What it is	Who controls the data
Billable hours	Time recorded and billed	The firm
Leverage	Work done by associates, paralegals and assistants	The firm
Realization and collections	The share of billed time clients actually pay	Clients decide, through discounts and write-downs
Pricing	Hourly rates, fixed fees, success fees	The firm, within what clients accept
Risk	Malpractice, sanctions, reputation	Courts, bars and insurers can see it

AI touches leverage first. The drafting, reviewing and summarizing that juniors do is the work AI does best. One study estimated 57% of lawyers' hourly work and 81% of assistants' could be automated. That matters twice: it cuts the hours a firm can bill, and it changes how juniors learn the work.

Next: Where AI touches the economics ›

Section 2 of 9

Where AI touches the economics

In one sentence

AI touches all five pools. In most industries, saved time is saved cost. In a firm that bills by the hour, saved time is lost revenue until pricing changes.

Pool	What AI changes	Size and timing	Cost or capacity?
Billable hours	Drafting, research and review take less time	Lawyers expect four hours a week saved within a year	Lost revenue under hourly billing; margin under fixed fees
Leverage	Junior work shrinks	Up to 57% of lawyers' hourly work exposed	Cost, if the firm hires fewer juniors; capacity otherwise
Realization	Clients push back on hours AI should have saved	Growing as clients learn what AI can do	Lost revenue, unless the firm prices ahead of it
Pricing	Fixed fees become more profitable as delivery cost falls	Flat-fee work up 34% since 2016	Margin, if priced on the new cost
Risk	Faster drafts, and new ways to be wrong	About 1,600 court filings with AI errors logged	Cost, if a mistake reaches a court

Capacity is not cost, and hours are not revenue. A firm that finishes faster has more capacity. It earns more only if it takes on more work, prices by the matter, or needs fewer people.

Survey figures are estimates. The time-saved and exposure numbers above come from companies that sell software to lawyers, measured their own way.

Next: What is already happening ›

Section 3 of 9

What is already happening

In one sentence

Most firms use some AI, few use it everywhere, and the results split two ways: firms that changed pricing are growing, and lawyers who stopped checking are being fined.

Who	What they did	Reported result	What the source does not show
US legal professionals (Clio, 2024)	Surveyed its users on AI	79% use AI, up from 19%; 8% use it everywhere; up to 74% of hourly tasks automatable	Run by a software vendor, among its own users
Growing and shrinking firms (Clio, 2025)	Compared firms by revenue trend	Growing firms nearly doubled revenue on 50% more matters and used AI more	Whether AI caused the growth
Legal and tax professionals (Thomson Reuters, 2024)	Asked how much time AI would save	Four hours a week in a year; twelve in five years	Expectations, not measured results
Courts worldwide (public database, 2026)	Logged filings with AI-made-up citations	About 1,600 cases, 1,116 in the US	Crowd-sourced, so a floor, not a full count
The national bar association (Formal Opinion 512, 2024)	Set out lawyers' duties when using AI	Verify output, protect client information, bill only time spent	Advisory; states adopt their own rules

Two patterns stand out. The firms gaining from AI are those that changed how they price, so faster work earns more instead of less. And the firms losing from it are those that trusted a draft they did not check.

Next: What could change next ›

Section 4 of 9

What could change next

In one sentence

Three shifts are under way: clients refusing to pay for time AI should save, the junior pyramid thinning, and AI errors becoming malpractice claims. Each has a signal to watch.

Scenario 1: Clients stop paying for time AI should save. Large clients write AI into their outside counsel guidelines: no billing for tasks AI can do, or a discount on them. Assumes: clients learn what AI can do faster than firms reprice. Signal to watch: one of your top five clients asks about AI in a bill review or a new guideline. Last checked: September 2026. Bar opinions already say hourly clients pay only for time actually spent.

Scenario 2: The pyramid thins. Firms need fewer first-year lawyers to do review and drafting, and must find new ways to train the ones they hire. Assumes: AI keeps taking on junior work. Signal to watch: peer firms cut entering classes or change how juniors are trained, citing AI. Last checked: September 2026. One estimate puts 57% of lawyers' hourly work within reach of AI.

Scenario 3: The made-up citation becomes a malpractice claim. A client loses a matter over an AI error and sues, or an insurer adds AI questions to the renewal. Assumes: errors keep reaching courts. Signal to watch: your malpractice insurer asks how the firm checks AI work. Last checked: September 2026. About 1,600 filings with AI errors had been logged by mid-2026.

All three reward the same preparation: a pricing plan for matters AI speeds up, a rule that people check every citation, and a way to train juniors on judgment, not just volume.

Next: Who gains and who loses ›

Section 5 of 9

Who gains and who loses

In one sentence

Firms that price by the matter keep what AI saves. Firms that bill by the hour hand the savings to clients and risk shrinking. Clients gain either way.

Clients gain. Work gets faster and, in time, cheaper. Clients who ask the right questions capture most of the savings.

Firms that price by the matter gain. When a fixed fee covers a matter, every hour AI saves is margin. The firms growing fastest in one survey were also the ones using AI most.

Firms that bill only by the hour lose ground. Every hour AI saves is an hour not billed. Unless they take on more work, revenue falls as the tools improve.

Juniors face a harder start. The work that trained them is the work AI does first. Firms that redesign training keep their pipeline; firms that do not will have fewer experienced lawyers in ten years.

Small firms and solo lawyers can gain the most. AI lets a small firm take on work that once needed a team, and price it competitively. The risk is that small firms have the least time to check what AI writes.

Next: Where to look before you invest ›

Section 6 of 9

Where to look before you invest

In one sentence

Four places are worth a mid-sized firm's money now. Each has a prerequisite that decides whether it pays, and a common way it fails.

Where	Prerequisite	Cost range for a $20M to $500M firm	Most common reason it fails	Pitch you will hear
Legal research and drafting tools	A rule that a person checks every citation against the source	$100 to $500 per lawyer a month	One AI is used to check another AI's work	"Research in minutes"
Document and contract review	Matter types with enough volume to learn from	$50,000 to $300,000 a year	Billed by the hour, so the savings cut revenue	"Review 80% faster"
Fixed-fee pricing for the matters AI speeds up	The real cost to deliver each matter type	Mostly partners' time	Fee set before the new cost is known	"Price for value"
A firm AI policy and client consent process	An owner, a tool list and consent language	Mostly staff time	Written once, followed by nobody	"Be compliant"

The cost ranges are illustrative and will be replaced with reviewed figures. The pattern holds regardless: change the pricing before the savings arrive, and check every word that leaves the firm.

One test before any contract. Ask the vendor to run its tool on a closed matter and compare its output with what your lawyers produced, including every citation.

Keep the rules that do not change. Lawyers must understand the tools they use, verify what they produce, protect client information and bill only for time actually spent.

Next: Up and down the chain ›

Section 7 of 9

Up and down the chain

In one sentence

Most of what AI will do to a law firm arrives through its clients: what they expect to pay, what they now do themselves, and what they need advice on.

Your clients: companies in every industry. In-house teams are using AI to do work they once sent out, and to question the bills for work they still send. Read the guides for your clients' industries: consumer goods, manufacturing and insurance each describe where AI is changing their legal needs.

Your clients' boards. Directors now ask about AI risk, regulation and liability. That is new advisory work for firms that understand it.

The other side of the table. The course "What AI is doing to your advisors" shows how clients are being taught to buy legal work in an AI era. Firms should read it too.

A good answer

Read, summarized and priced into proposals.

Push back if you hear

“Nothing yet.”

Next: What changed ›

Section 9 of 9

What changed

In one sentence

No changes yet; this is the first version. Every dated change goes here, newest first, and any change that alters a conclusion is marked and the section updated that week.