In the previous blogs in this series, we established that most firms are running three GenAI strategies simultaneously, and that capital allocation has usually already sequenced them — even without an explicit decision.
The natural response from most managing partners at this point is reasonable: so what? Practice groups have always moved at different speeds. That’s how law firms work.
They’re right. And that’s precisely why GenAI creates a different challenge.
The Matrix That Already Exists
Walk through most large firms today and you’ll find something like this:
The M&A group adopted a premium drafting tool early, is seeing real efficiency gains, and is quietly absorbing the cost because it helps win work. The litigation group is still in pilot because the partners are skeptical and the technology hasn’t proven itself on complex discovery. The employment group is using a general-purpose AI tool informally, outside the approved stack, because the approved tools don’t fit their workflow. The real estate group is waiting for guidance from firm leadership before committing to anything.
Four practice groups. Four different positions. All rational given their specific client mix, partner composition, and competitive context.
This isn’t new. It’s how law firms have always absorbed new technology — unevenly, organically, at the pace of the most influential partners in each group.
So why is GenAI different?
Three Reasons the Old Logic Breaks Down
First, the client conversation is now firm-wide.
When a client’s GC sends an updated outside counsel guideline asking how the firm’s AI investments will benefit them, they send it to the relationship partner — not to a practice group. The answer that partner gives needs to be coherent across the firm.
A matrix of four different adoption postures produces four different answers, and sophisticated clients will notice the inconsistency.
In previous technology cycles, clients rarely asked. Now they’re asking directly, in writing, often with procurement teams involved.
Second, the cost structure is shared — even when licenses are seat-based.
Most GenAI platforms are purchased in bundles or minimum commitments. Even when priced per seat, they often require parallel investments: cloud upgrades, security configurations, data governance enhancements, device refresh cycles.
The financial exposure sits at the firm level, even if adoption decisions are made at the practice-group level.
Uneven adoption across practice groups doesn’t just create strategic inconsistency; it creates a capital allocation problem that sits on the CIO's or COO’s desk. When utilization is concentrated in a few groups while others lag, the ROI argument for the broader investment weakens.
That pressure lands centrally — even when the adoption decisions were made locally.
And the concern about subsidizing certain practice groups becomes real.
Third, the reputational exposure doesn’t stay local.
If a practice group uses AI in a way that produces a client-facing error — a hallucinated citation, a confidentiality breach, an incorrect summary — the reputational consequence isn’t contained to that group. It’s a firm story.
The governance infrastructure that protects against that risk has to be consistent across the firm, regardless of how each group has chosen to adopt.
You cannot have a Trust First strategic posture institutionally while allowing Competitive Positioning First behavior in individual practice groups without accepting real exposure.
What This Means for Strategy Sequencing
The goal isn’t uniformity.
Expecting the M&A group and the employment group to adopt at the same pace, using the same tools, with the same metrics, is neither realistic nor desirable. Practice groups should adapt to their clients' needs.
But there are three things that cannot be local decisions:
Governance standards — the floor of what is permissible, regardless of which tools a group uses or how advanced their adoption is. This protects the firm, not just the group.
The client narrative — the answer that any partner in any group can give when a client asks about AI. It doesn’t need to describe every tool in use, but it needs to be consistent and defensible. One firm, one story.
The investment rationale — the basis on which enterprise-level spending gets defended, adjusted, or expanded. That decision has to be made with a view of the whole firm, not just the groups that are most enthusiastic or most resistant.
Everything else can flex. Which tools each group uses, the pace of rollout, the specific use cases prioritized — these can and should vary.
The matrix is fine.
But the matrix needs a frame.
The Conversation That’s Needed
The conversation that is needed isn’t a strategy retreat conversation. It’s shorter and more specific, and it needs to happen between the managing partner, CIO, COO, and the two or three practice group leaders whose adoption decisions are setting the de facto pace for everyone else.
The questions aren’t “what should our strategy be?” They’re:
- What are the firm-wide non-negotiables, regardless of how quickly each group adopts?
- What guidance does central leadership need to provide so local adoption decisions align rather than conflict?
- If a client asks any partner in any group about our AI posture tomorrow, what do we want them to say?
The output isn’t a strategy document.
It’s a short, shared answer to that last question — and clarity on the governance standards and investment logic that apply firm-wide, even as adoption speeds differ.
That’s achievable.
And it’s the minimum required for law firm AI strategy alignment — without forcing the firm to resolve every tension at once.