Legal bills from outside law firms are some of the biggest checks companies write each year. Artificial intelligence is suddenly giving in-house lawyers a way to shrink those costs.
When Doug Mandell, the general counsel of chatbot-maker Inflection AI, was working on a new data security policy, he turned to a tool from the legal AI startup GC AI. It chewed through his notes and background material and spit out a draft policy. Mandell then workshopped it in Google Docs before sending it to an outside law firm for review.
Before he had the tool, Mandell said, he would email outside counsel saying, “We’re thinking about doing X, Y, Z,” and then wait, racking up billable hours as the law firms tinkered. Now, he tells them, “Start with this.”
“To be clear, we’re not replacing outside counsel,” Mandell said. “We’re refining the process in a way that benefits the company.”
AI has landed in law, but the transformation hasn’t been evenly distributed. Tools that make lawyers faster are a difficult sell for firms that bill by the hour. The more aggressive adoption is coming from in-house legal teams, where efficiency isn’t a threat, but a mandate.
In-house lawyers don’t get paid more for taking longer. They’re often judged on how little they slow the business down, because delays can mean a deal slips or a job candidate walks.
“In-house lawyers know a finger is being pointed at them, and nobody wants to be in that situation,” said the venture capitalist Bogomil Balkansky, whose Silicon Valley firm Sequoia Capital has backed legal startups including Ironclad, Harvey, Crosby, and Sandstone.
That dynamic, he said, helps explain why corporate legal teams are becoming eager adopters of new technology that cuts a task from days to minutes, or helps keep work in-house. It offers leverage to move faster — and to shed the reputation of being the department that says no.
Similar to Mandell, many in-house lawyers now have access to tools that draft legal documents, check compliance, and compare terms across agreements. Software can read contracts line by line and adjust terms to match a company’s internal rules and preferred negotiating positions.
That demand is reshaping how legal tech companies build and sell their products.
Harvey, a startup that began serving major law firms, has pushed into large enterprises, signing up legal teams at Walmart, General Mills, and Bayer, among others, last year.
John Haddock, Harvey’s chief business officer, said a general counsel’s first priority is speeding up contracting and deflecting questions they get from colleagues on the business side. The goal isn’t to replace outside counsel, he said, but to make in-house teams more productive at the work they already do. It’s relief from the “high-volume, high-toil” work that eats up scarce time.
Though, some billable hours are disappearing, said Cecilia Ziniti, a former lawyer at Amazon and Replit who now runs GC AI, the service that Mandell uses. In a recent survey of 100 in-house lawyer users, 14% reported savings in outside legal spending after adopting the tool. Ziniti said GC AI grew annual revenue from $1 million to $13 million last year.
AI is already embedded in the daily legal work at companies like Gusto.
Dina Segal, the company’s chief legal officer, oversees a team of 70 lawyers who use AI to monitor thousands of rules across employment law, benefits, and corporate compliance. Their system is trained on Gusto’s business, delivering rule-change summaries through Slack and email and routing them to the right lawyers. Segal said it’s been a “game changer,” letting the company’s legal experts spend less time tracking rules and more time advising the business.
When Gusto does engage outside counsel, Segal said, it’s typically for bet-the-company matters — complex, nuanced strategic decisions. And even then, the conversations don’t start from scratch. Gusto’s lawyers are caught up on the regulatory changes for a policy discussion, for example.
“You can come in with an early perspective instead of a blank piece of paper,” Segal said, “then build from there together.”
As in-house legal teams move faster, they’re quietly resetting expectations for the law firms they hire. Haddock said he increasingly hears from general counsels who have surveyed their outside firms about how they’re using AI and are factoring the answers into hiring decisions.
That shift is happening against a backdrop of stagnant budgets. In a Q3 Thomson Reuters survey of general counsels at global companies with more than $1 billion in revenue, 35% said they planned to increase legal spending in the year ahead, down from 40% at the start of 2025.
“The crucial insight,” the report notes, “isn’t that there may be less legal work on the horizon,” but that clients are being forced to make “increasingly brutal choices about which firms get their limited dollars.”
Mandell, the Inflection AI lawyer, is clear about where the line is. He said he wouldn’t pay a law firm $900 an hour to hunt through documents for answers he can get instantly from a tech tool.
He’s paying for judgment. “When my back is up against the wall, and I have to make a decision,” he said, “I want to talk about it with a human being.”
Have a tip? Contact this reporter via email at mrussell@businessinsider.com or Signal at @MeliaRussell.01. Use a personal email address and a non-work device.
