In-house AI is a shift of legal work away from outside counsel

On October 20, Legaltech News (Law.com) reports that legal departments are cutting down on the work they send to outside counsel. The report says law firms are passing on AI-driven savings and looking at new billing models. That is the news: in-house use of AI is being associated with less external legal work, while outside firms are adjusting how they charge for what remains.

The development challenges the idea that in-house teams may keep some outsourced work without losing share of the market. Legal departments do not need to eliminate outside counsel for that position to weaken. Sending less work externally is enough to put pressure on the outside share, particularly when the same development also pushes firms to pass on AI savings and reconsider billing. General Legal describes the commercial tension more directly, saying that 71% of in-house legal professionals expect firms to change their commercial models as AI adoption grows, while 62% of law firm professionals say pricing remains unchanged. Its conclusion is that clients see the disconnect between billing for time and AI saving it. The direction is therefore not simply that in-house teams are becoming more capable; it is that capability is changing where work is performed and how the remaining work is priced.

What this establishes is narrower than a forecast of general law-firm decline. Legaltech News reports a reduction in work sent to outside counsel, but it gives no amount for that reduction, no market-share figure, and no indication that in-house teams are keeping every matter they bring inside. It also does not show whether outside firms are replacing lost volume with new work, or whether new billing models preserve revenue while changing the unit of sale. General Legal supplies survey percentages about expected commercial change and current pricing, not a measured transfer of matters from firms to legal departments. The stronger reading would require observable movement in outsourced work and in the outside share of legal spending, alongside evidence that in-house teams are retaining work rather than merely using AI to handle a larger portion of the same workload. Those conditions are not established here. The event does establish a direct pressure point: when internal teams do more with AI, outside counsel cannot assume that the old flow of work remains intact.

The next confirming fact would be a law-firm or legal-department disclosure showing that a defined class of matters moved inside, with the volume or share of outsourced work stated before and after the change. A second confirming fact would be a disclosed fixed-fee or other commercial model tied to that shift, showing whether firms are responding to lost external work by changing price rather than merely changing invoices. A result cutting against the position would be an outside firm reporting that AI reduced the labor required on recurring matters while the volume and revenue of those matters held, without billing for unworked time. We will treat the direction as established only when the next announcement, filing, or client survey puts numbers on both sides of the boundary: work retained in-house and work still purchased from outside counsel.