The Great Unbundling: Why Corporate Legal Departments Are Bringing More Work In-House

The relationship between corporate legal departments and outside counsel has always been transactional, but it has never been symmetrical. For decades, law firms held the capability advantage: the institutional knowledge, the bench depth, the specialized expertise that even well-resourced GCs couldn't replicate internally. That asymmetry is dissolving—and faster than most BigLaw managing partners are willing to admit.

The data is unambiguous. According to the ACC/Everlaw 2025 survey of 657 in-house professionals across 30 countries, 64% of in-house counsel now expect GenAI to reduce their reliance on outside firms—up from 58% just a year earlier. Simultaneously, the Harbor/CLOC 2025 Law Department Survey of 135 departments with a median revenue of $13 billion found that only 37% of legal departments anticipate increases in outside counsel spend—down sharply from 58% the prior year. These aren't the numbers of a cost-cutting exercise. They're the numbers of a structural power shift.

The Numbers Behind the Shift: How Much Work Is Actually Moving In-House

The insourcing movement predates generative AI, but AI has given it decisive momentum. By 2024, in-house lawyer counts matched those at all top-500 law firms combined—a milestone that rarely gets the attention it deserves. Citrin Cooperman's analysis found that 66% of companies had already planned to bring legal work in-house and that one in four legal departments was actively reducing its law firm roster.

The CLOC 2026 State of the Industry Report confirms the operational logic driving this: legal demand is surging—regulatory complexity (cited by 63% of departments), cybersecurity governance (58%), and contract volume (53%)—but budget and headcount growth have flatlined. Only 32% of departments expect attorney headcount increases. The math forces a productivity solution, and AI is that solution.

The Harbor survey found that 65% of departments made intentional efforts to insource work within the past one to two years, while 85% now have dedicated resources or committees managing AI adoption—a decisive leap from experimentation to enterprise integration. This isn't tinkering; this is infrastructure buildout.

AI as the Equalizer: How Technology Is Closing the Capability Gap

For most of legal history, the capability gap between a sophisticated in-house team and a large firm was real and large. A GC could hire smart lawyers, but she couldn't replicate the cross-client pattern recognition, the deep specialist bench, or the institutional memory that came from advising hundreds of similar matters. AI doesn't eliminate that gap entirely—but it compresses it enough to matter.

In-house GenAI adoption has more than doubled in a single year: 52% of in-house teams actively used GenAI in 2025, up from 23% in 2024. Notably, in-house teams have actually outpaced law firms on AI tool adoption, with 34% of in-house teams using AI-specific tools versus 19% at midsize firms.

The GC dissatisfaction with law firm AI adoption is pointed: only 24% of in-house counsel feel satisfied with how outside counsel are adopting GenAI for cost savings, and 59% report receiving no GenAI-related savings from their law firms at all. Firms are investing heavily in AI—technology spending jumped 9.7-10.5% across the industry in 2025—but they're not passing efficiency gains to clients. That gap in expectation versus delivery is accelerating the insourcing decision.

Contract Management, Discovery, and Due Diligence: The First Work to Go

The ACC data is specific about which work is migrating first. 78% of in-house teams plan to insource drafting work, 71% contract management, and 62% research—precisely the high-volume, process-driven work that has historically funded mid-tier law firm revenues. These are not incidental practice areas; they represent the billable foundation of hundreds of firms.

Contract lifecycle management tools—Kira, Luminance, SpotDraft—give in-house teams the ability to review, extract, compare, and flag clause-level risk across thousands of agreements without external support. AI-powered e-discovery platforms have similarly commoditized document review. And as Bain & Company research cited by M&A advisors notes, 21% of M&A professionals were already using generative AI in transaction processes as of 2025—with near-universal adoption expected within five years.

The implication is direct: the due diligence factories that BigLaw and mid-tier firms operated, often staffed by junior associates billing at $500-$700/hour for document review, are structurally uncompetitive against AI-enabled in-house teams running the same process for a fraction of the cost.

What GCs Say About Why They're Pulling Back (It's Not Just the Rates)

The conventional narrative frames insourcing as a budget response. The actual GC complaint is more nuanced—and more damaging to firm positioning. All surveyed GCs in Citrin Cooperman's analysis expressed concerns with cost, quality, or other challenges—meaning rate is only one variable in the dissatisfaction equation.

Responsiveness is the underreported driver. GCs report inconsistent service standards, where outside counsel engage differently with company leadership versus operational staff. The value proposition of a firm that is expensive and slow and not deploying AI efficiency gains back to clients is difficult to defend. 61% of in-house survey respondents intend to push for change in how legal services are delivered and priced—a figure that signals not passive withdrawal but active pressure campaigns.

Meanwhile, Gartner's December 2025 survey found that only 20% of legal matters sent to outside counsel stay within the budget range—a budget predictability failure rate that no CFO would tolerate in any other vendor relationship. When law firms lose the cost predictability argument alongside the efficiency argument, the case for retaining them on routine work collapses.

The Firms Most at Risk—And Which Practice Areas Face Structural Obsolescence

Not all firms face the same exposure. The 2026 US Legal Market analysis already reveals a telling bifurcation: midsize firms grew demand nearly 5% in the latter half of 2025 as GCs redirected routine work to lower-rate providers, while Am Law 100 firms couldn't exceed 2% demand growth despite concentrating on premium matters. The squeeze is hitting the middle most severely.

Firms whose revenue depends on high-volume, process-oriented practice areas—standard commercial contract work, routine corporate governance, predictable compliance filings, first-pass document review—face structural, not cyclical, obsolescence. These practices will not recover their prior volumes when AI adoption matures further. The ALSP market, already at an estimated $28.5 billion and growing at 18% annually, is capturing additional share that previously flowed to law firms.

The firms best positioned for the unbundled era are those whose work genuinely cannot be replicated internally: bespoke deal structuring in complex cross-border transactions, high-stakes litigation requiring trial experience and courtroom relationships, regulatory strategy in novel enforcement environments, and crisis management requiring C-suite credibility. These practice areas will command premium rates precisely because their complexity and consequence make internal replication impractical.

The Survival Playbook: How Outside Counsel Can Remain Indispensable

Firms that treat this as a pricing negotiation will lose. The correct strategic response is repositioning: from generalist service provider to irreplaceable specialist, from staff augmentation to genuine strategic counsel.

This means three concrete moves. First, firms must explicitly define and defend the work only they can do—cross-client precedent depth, trial team experience, regulatory relationship capital, crisis credibility—and stop competing for commoditizable volume where they cannot win on cost. Second, firms must demonstrate genuine AI-enabled cost pass-through; the 59% of GCs receiving no AI savings from their firms represent a service expectation that will force convergence or defection. Third, firms should position embedded expertise as a service model: secondments, subscription-based advisory arrangements, and fractional GC structures that deliver institutional knowledge without the full hourly engagement model.

The power dynamic between GC and outside counsel is rebalancing permanently. In-house teams are not building capabilities to replace firms on bet-the-company work—they're building capabilities to stop paying law firm rates for everything else. The firms that acknowledge this distinction and specialize accordingly will retain indispensable status. The firms that don't will find their client base eroding not from a single dramatic client departure, but from a hundred incremental insourcing decisions made quietly over the next three years.