Over the past two years, most law firms have made real progress on GenAI. Tools approved. Policies written. Training launched. By any reasonable measure, firms have been moving.
And yet something keeps surfacing in leadership conversations: a sense that the investments aren't adding up to a direction.
In my conversations with managing partners, CIOs, and COOs, what I observe most often isn't inaction — it's three legitimate strategies running simultaneously, none of which has been made explicit. And that's where the friction comes from.
The Three Strategies Already Running in Your Firm
Strategy 1: Trust First
Risk, compliance, and senior client-facing partners are focused on protecting reputation above all else. Careful governance, approved tools, controlled rollout, no surprises. Their measure of success is: nothing goes wrong.
Strategy 2: Capital Discipline First
The CIO, COO, and finance leadership are watching a cost base that grew nearly 10% in technology spending in 2025 — without clear utilization metrics to justify it. Their measure of success is: every dollar earns a return.
Strategy 3: Competitive Positioning First
Business development leaders and practice group heads in competitive markets see peer firms signaling AI capability to clients and recruits. Falling behind on perception has real costs. Their measure of success is: clients and talent see us as forward-looking.
None of these is wrong. All three are rational responses to real pressures.
The problem is that when all three operate implicitly and simultaneously, every significant decision becomes an unspoken negotiation. Tool approvals stall. Pilots run indefinitely. The loudest voice in any given room wins — and that voice changes depending on who's in it.
Why Sequencing Matters
A firm ultimately needs all three. But they cannot all lead at once.
| Trust First | Capital Discipline First | Competitive Positioning First | |
| Strength | Low headline risk, strong governance narrative | Budget resilience, clear ROI justification | Strong recruitment story, visible client differentiation |
| Risk | Can appear less ambitious than peers | External positioning may lag while you build internal depth | Signal must convert to substance — or the strategy becomes fragile |
| Trade-off | You cannot govern and lead simultaneously — committing to trust first means accepting that some competitive ground gets ceded, at least initially | You cannot build deep utilization and broad capability at the same time — committing to discipline means accepting a slower external story | You cannot signal boldly and spend conservatively — committing to positioning first means accepting a higher cost base and the internal pressure that follows when utilization doesn't immediately match investment |
| Client experience | Reliable, conservative, fewer surprises | Increasing efficiency over time | Access to premium tools, innovation messaging |
The order shapes everything: what gets funded, what gets measured, and critically — what partners can say when a client asks how your AI investments benefit them.
Why This Matters More Now Than It Did in 2025
In a strong market, strategic ambiguity is affordable. When revenue is growing, all three strategies can coexist without forcing a hard conversation — the trust faction governs, the discipline faction tracks utilization, the positioning faction signals, and nobody has to reconcile the tensions because there's enough budget to fund all of it loosely.
According to a recent Thomson Reuters report, that window is closing. Client spend anticipation among corporate legal departments has dropped to its lowest point since the pandemic. Only 13% of buyers plan to increase external legal spend. Clients are simultaneously building their own AI capabilities and scrutinizing what they're paying external counsel for.
The practical test isn't abstract: imagine being asked to find 10-15% in cost reductions, or facing a meaningful revenue contraction. Which investments would you defend, and which would you cut? If that question is hard to answer — if the answer changes depending on who's in the room — that's a signal that the strategy is still implicit.
Firms that have sequenced deliberately will have options. They'll know what to protect and what to adjust. Firms that haven't will find themselves making those decisions reactively, under pressure, with clients watching.
That's not a prediction. It's a planning question every managing partner, CTO, and COO should be able to answer today.
That question is already arriving in a different form — updated outside counsel guidelines asking how AI efficiency gains will benefit clients, or stating that firms' tool costs won't be covered. The firms with a clear GenAI strategy will handle those conversations from a position of confidence. The firms without one will improvise.
The Practical Implication
In innovation management, opportunistic exploration is natural. Early pilots, parallel experiments, and uneven adoption are often how firms learn. The challenge isn’t that this happened. The challenge is failing to transition from opportunistic exploration to deliberate direction once investments scale.
Most firms haven't sequenced explicitly because doing so requires naming the tension between factions that all have legitimate standing. But avoiding the conversation doesn't resolve the tension — it just means someone else resolves it for you.
The goal isn't perfect consensus. In a partnership, that's rarely achievable. The goal is a working direction that leadership can defend, partners can carry into client conversations, and the CIO or COO can use to evaluate whether current spending is pointed in the right direction.
In the next piece, we'll look at how to have that conversation — starting not with a blank whiteboard, but with the investment decisions your firm has already made. Because in most firms, the capital allocation has quietly answered the strategic question. The real conversation is whether that answer was intentional — and whether it is the GenAI strategy the firm wants to sustain if 2026 turns out to be less prosperous than 2025.