Corporate legal departments are moving faster than their outside counsel on generative AI, positioning themselves to take on more work internally and apply new pressure to long-standing law firm relationships, according to a new survey of legal operations leaders.

The findings come from the Blickstein Group and FTI Consulting’s 18th annual Law Department Operations Survey Report, covered by Legaltech News. Based on responses from 68 mostly North American companies, the survey highlights a widening adoption gap between clients and their firms at a time when legal departments face growing pressure to deliver faster service at lower cost.

In-house teams reported substantially higher gen AI usage across core legal workflows. For legal research, 72.3% of respondents said their departments use gen AI, compared with 23.4% at outside firms. Similar gaps emerged in drafting, document review, and contract-related work. Even in areas where adoption is closer, such as e-discovery, corporate legal teams still reported higher usage than their counsel.

The implication is structural. Nearly 94% of respondents said gen AI will allow them to bring more legal work in-house, an increase from already elevated expectations last year. At the same time, 87.5% expect to spend less on law firms, a jump of more than 20 percentage points year over year.

Despite that outlook, most legal departments have not yet asked firms to demonstrate AI-driven cost savings. Fewer than one in five respondents reported pressing outside counsel on that point, a hesitation that may reflect uncertainty about how savings should be measured in an AI-enabled legal workflow. Brad Blickstein, CEO of the Blickstein Group, told Legaltech News that many law departments lack a clear strategy for what they want firms to deliver using AI.

Law firms, however, may be misreading the calm. Meredith Brown, a senior managing director at FTI Consulting, said legal departments are actively identifying high-volume areas—such as contracting and privacy work—that may increasingly be built in-house rather than sent to firms. The shift is driven less by innovation narratives than by volume, speed, and cost control.

The data also complicates assumptions about alternative legal service providers. Nearly two-thirds of respondents said their ALSP spending declined over the past year, with half expecting further decreases. When selecting ALSPs, use of advanced technology ranked behind price, expertise, and willingness to adopt alternative fee arrangements.

Inside legal departments, the priority is not narrowly framed as cost cutting. More than 90% of respondents cited efficiency and performance improvements as the primary drivers of AI strategy, compared with 57.7% who pointed to cost savings. While nearly two-thirds said they could demonstrate non-financial benefits from gen AI, only about one-third were able to show measurable savings.

Instead, legal leaders are pointing to throughput. As Brown put it, departments are increasingly doing more work with the same budgets, tackling matters that previously would have gone unaddressed or been outsourced.

Most legal ops teams believe they have access to the right technology today, but control over future tools is uneven. Fewer than 40% said they have primary responsibility for selecting AI-powered legal assistants, suggesting enterprise procurement decisions may shape legal AI adoption as much as legal strategy.