New Research: Concerns Aside, AI is Delivering Results for In-House Legal
Corporate in-house teams, seeing efficiencies internally and from their outside counsel, prepare a new approach to billing
The vast majority of corporate legal teams say they – and their outside firms – have embraced generative AI and are gaining meaningful advantages from the technology. While concerns around accuracy and other risks remain, in-house professionals say the technology will continue to drive change, with new approaches to the billable hour on the near horizon.
In a survey report of 300 in-house legal professionals, released today by Everlaw in partnership with Wakefield Research, 86% of respondents said they are using generative AI tools for legal work. Perhaps more importantly, 92% say they’re seeing measurable cost savings, in part because they’re able to handle more work internally, rather than handing it off to outside counsel.
“AI tools are making in-house teams more efficient and productive,” says Everlaw Chief Legal Officer Gloria Lee. “But I also love that it's helping us do things that we really could not do before, like using AI tools to harvest inventions for patent disclosures.”
The research also captures how corporate teams view their outside firms going forward: perhaps less as a way to scale up for rote labor, and more as a source of specialized expertise. Another change in perspective involves how legal services are priced in the age of AI. Respondents widely predict a rapid end to the primacy of the billable hour.
These and other trends outlined in the In-House Innovation Report: Corporate Teams Gain ROI and Leverage all suggest not only that change is coming faster than we expect, but that much of it is already here.
AI Gives In-House Legal More Control Over Outsourcing
In a key finding, 54% of respondents reported that, thanks to the capabilities and efficiencies of AI, they’ve already reduced the volume of work they’re sending to outside counsel. Another 44% said that they intended to start cutting back on outsourcing within the next year. In other words, nearly everyone will be delegating less work to outside firms in 2027.
This does not mean that outside firms will have nothing to do. In fact, Lee said, the decline in the volume of work will largely be in the lowest-value tasks, the ones where clients need outside counsel’s sheer labor more than their special expertise. This, she said, elevates the value that outside firms bring.
“AI rebalances the relationship between clients and their law firms,” Lee says. “As in-house teams take more control of certain types of work, our law firms can act as true strategic advisors on our highest-value challenges.”
At the same time, as the research shows, a relationship centered more on the value of expert insight than the volume of labor has significant implications for the billable hour.
Law Firms Are Passing Along AI Savings
A year ago, in-house legal professionals seemed to have little idea whether their outside firms were even using GenAI: In Everlaw’s 2025 research, nearly three-fifths said they didn’t even know whether their law firms were using AI at all.
That has definitely changed in 2026. Not only do corporate legal teams know that their firms are using GenAI, only 8% say they haven’t experienced savings as a result.
13% of respondents say they’ve seen “significant” GenAI-related cost savings
56% say their savings have been “moderate”
23% say outside counsel savings have been “slight”
This increased transparency is a positive result, and one that should help outside firms and their clients navigate the changes in legal work driven by AI and automation.
Change Is on the Horizon: Billable Hour vs. Billable Value
The billable hour is central to law firms’ business model, but also a great challenge for in-house teams tasked with controlling costs.
Tensions around time-based billing and the slow growth of alternative fee models have bubbled through the industry for decades. Now, with GenAI-based tools showing increasing ability to compress hours or weeks of human work into near-instant outputs, in-house professionals expect value-based fees to soon eclipse the billable hour.
Eighty-one percent of respondents say that within five years, most legal work will be priced by outcome rather than billed by hours worked, and 47% say it’ll happen in only three years. Only 1% of respondents think that shift will take more than a decade, and no one said it won’t happen at all.
The ingredients of such a change are already evident. As corporate teams reduce the amount of work they send out, and demand greater transparency into outside firms’ AI efficiencies, a discussion around hours billed seems inevitable. And while law firms might be expected to resist change to their core business model, it might actually make sense for them to base their fees on the high value of work that might’ve been done more quickly thanks to AI tools.
Everlaw’s Lee says that both sides are still dancing around the topic, with neither taking decisive action. “Clients have the ability to drive the shift to outcome-based billing,” Lee says. “They have the motivation, which is considerable cost savings, and they have the power, because already they’re able to reduce their dependency on outside firms."
Adoption Is High, But Concerns Remain
From the beginning of the GenAI era, wonder at an LLM’s way with words has been tempered by concerns at its ability to confidently make things up. While models have improved in the past four years, errors and hallucinations still happen.
Respondents’ top concern about AI, indicated by 80% of them, is accuracy and alignment. A close second (72%) is that non-lawyers will rely on AI for legal advice that may also be error-prone.
And for good reason: a third of respondents say they’ve seen fabricated evidence or non-existent case law show up in litigation. Seventeen percent say they’ve seen AI-related errors from opposing counsel, 12% have seen them from their own team, and 7% from a witness, all of which underscores the need for human oversight of AI tools.
Another concern is skill degradation. Research into the use of AI generally suggests that users become reliant on the tools and less able to perform effectively without them. Close behind immediate concerns about accuracy is the worry, cited by 63% of respondents, that overreliance on GenAI will erode human skills.
Given that AI tools are fast evolving and professionals are still learning how to use them effectively, it makes sense that corporate legal teams are prioritizing upskilling and continued human oversight of AI in the year ahead, with 68% saying they prioritize staff training and 53% will be focusing on internal governance, compliance, and policies for AI usage.
Legal Pros Lean Into the AI Future
Past research has emphasized that GenAI will change legal work. This year’s survey indicates that the change is definitely here. Corporate teams are both reducing the amount of rote, manual labor they must perform, and the amount of it they must send to outside counsel. The result is greater speed and early cost savings.
Yet more change must come. Not only will the relationships between law firms and their corporate clients have to evolve, but legal professionals of all stripes will determine where automation and generative tools are most useful, what skills are required for this new era, and how to maintain accuracy and accountability for machine-generated outputs.
One thing our research has shown in the last few years is that, contrary to its traditional role as a tech laggard, the legal industry is up for this challenge.
Download your free copy of the In-House Innovation Report: Corporate Teams Gain ROI and Leverage and see how your peers are preparing.
