This Week’s Theme
Law firms are talking about AI as if the technology were the strategy. It isn’t.
And because the AI hype is causing modernization without a foundation, they chase tools, copy what others are doing and call it innovation.
But not a lot changes because the underlying structure; how the firm works, decides, measures, and prioritizes; has never been built.
AI doesn’t fix that, it exposes it.
The pattern extends across the field:
-
no clarity about what the firm is actually trying to become
-
no definition of the problem the firm exists to solve
-
no distinction between work that creates value and work kept alive purely by tradition
-
no grasp of the competencies the next decade demands but the current culture suppresses
In that environment, AI becomes noise. And every decision becomes a reaction, not a strategy.
This week is about why firms copying each other will hit a ceiling they can’t grow past.
You’re Not Behind on AI, You’re Behind on strategy
A structural truth is becoming unavoidable: AI on top of a missing strategic layer ends in a failed attempt.
Several behaviors keep repeating:
1. Copy-paste decision-making
Most firms don’t want to be first. If no one else is doing it yet, they avoid it.
2. Starting with tools and not introspection
Buying AI without defining the problem only locks the firms into expensive dead ends.
3. Efficiency over effectiveness
Making a broken process faster just multiplies the damage.
4. Misaligned incentives
Individual targets kill collaboration and make transformation impossible.
5. External pressure vs internal capability
Clients are evolving faster than the firms trying to serve them.
Legal AI in Action
🎬 The MCP Access Problem
Your AI assistant now has a broader attack surface than your staff, and no one is governing it.
🎬 The AI Risk Legal Teams Aren’t Testing For
Autonomous systems acting under your identity, with no supervision, no logs, and full access to client data.
Red Flag of the Week
Most AI vendors promise zero-retention and strong privacy. But those promises only hold if the company stays healthy.
If the vendor needs new funding, restructures debt, or gets acquired, your privacy terms can be rewritten overnight.
The danger isn’t the model. It’s the vendor’s ability to survive long enough to honor the contract.
What the Legal Frontlines are Saying
Based on my discussion with law firms’ IT leaders this week, three clear patterns keep repeating:
1. Firms treat AI like normal software
Most checks stop at a security questionnaire. Very few firms look at the real risks: data exposure, governance gaps, or the cost of a breach (average $5.5M for shadow AI).
2. Leadership still assumes “it won’t happen to us”
But IT managers stress the same message: it’s not a question of if, it’s when. AI increases the risk because it connects into email, drives, and case systems the firms already depend on.
3. The biggest gap isn’t tools, but ownership
Most firms don’t know who is responsible for AI decisions. But when something goes wrong, it’s the partners and not the office manager who get named, fined, and held accountable.
Looking Ahead
🎙 This Saturday at 2pm CET!
This week’s guest on Rok’s Legal AI Conversations: Lev Loukhton, former Sullivan & Cromwell and Linklaters M&A partner turned legal tech investor. We talk about client-driven adoption, the pressure on traditional business models, and what managing partners need to fix before investing in AI.
Each edition of Legal AI Brief brings practical lessons from firms using AI safely.