Law Punx: "More Strategic Work" Is An AI Myth
There's a comforting story spreading through law firms: AI takes the drudge work, and lawyers finally get to do "more strategic work." It sounds neat. It's also mostly false.
Many lawyers won't get that shift. And even for those who do, it's high-stress, high-stakes work that not everyone actually wants all day. On top of that, AI will keep climbing the value ladder. There's no guarantee it will stop at basic tasks.
The myth in plain terms
"More strategy" doesn't magically appear because admin tasks drop. Strategy is constrained by client demand, risk appetite, and who controls the relationship. Firm economics still favor leverage, repeatable matters, and predictable margin.
Partners won't turn every associate into a strategist. In-house teams won't expand strategic headcount just because drafting got faster. The market decides, not the tech.
Yes, AI will move up the stack
AI is already handling research, clause swaps, and first drafts. It will improve at pattern spotting, issue trees, scenario analysis, and triage. That's the path.
High-level legal judgment won't vanish, but the surface area of "strategic" tasks that require a human will shrink. That's a feature of progress, not a bug. Even major consultancies expect sizable automation of knowledge work over time. Source.
So what actually changes for legal teams
- Hours lose value. Outcomes, risk reduction, and speed become the product.
- Work splits into two tracks: systems and judgment. Most people will live in the systems track.
- Leverage shifts from people to playbooks, data, and AI-enabled workflows.
If you're in private practice
- Rebuild pricing: move repetitive work to fixed fee or subscription with clear SLAs.
- Treat your clause libraries, prompts, and QA checks as firm IP. Maintain them like products.
- Make "explainability" part of delivery. Show clients how you verified AI output.
- Measure and publish turnaround times, defect rates, and escalation thresholds.
If you're in-house
- Define decision rights: what AI can auto-approve, what requires legal review, what escalates.
- Turn policies into executable checklists a system can follow and audit.
- Stand up vendor and model risk reviews. Contracts, privacy, retention, bias checks-repeatable and logged.
- Educate business teams on what AI can and can't do today. Reduce "drive-by" legal asks with templates and guardrails.
Practical workflow to adopt now
- Scoping: convert matter intake into structured instructions (facts, constraints, risk posture, deadlines).
- Generation: let AI produce drafts, issue lists, and negotiation positions aligned to playbooks.
- Validation: run automated checks (policy compliance, deal math, clause conflicts), then human review on exceptions.
- Decision: log the rationale for key calls. Keep an audit trail that a regulator or court would respect.
Skills that compound for lawyers
- Framing problems so AI hits the target (clear instructions, constraints, examples).
- Quality control: testing prompts, creating checks, spotting failure modes fast.
- Negotiation frameworks: what to concede, what to trade, what never moves.
- Client education: simple explanations of risk and options, not jargon.
- Process design: turning legal judgment into steps a team and a system can run.
Career paths that make sense
- Strategist: rare spike matters, risk calls, board-level advice. Fewer roles, higher stress.
- Builder: playbooks, templates, data, tooling, and reporting. Multiplier impact.
- Operator: high-throughput matters with strong QA. Reliable, scalable, valuable.
- Relationship owner: trust, communication, and outcomes alignment with the business.
Billing and value, rethought
- Price insight and risk transfer, not time. Offer options: fixed, subscription, or outcome-based.
- Publish what "good" looks like: response times, accuracy targets, escalation rules.
- Bundle training and templates. Clients want fewer touchpoints and more certainty.
Reality check
AI won't hand you a calendar full of strategy. It will compress tasks, expose weak process, and reward teams who productize their best thinking. That's your opening.
Build systems. Keep judgment for the calls that truly move risk. Let everything else get faster.
Context
These points echo a recent Law Punx talk that questioned the "more strategy" story and pushed for a more honest plan: retool work, reshape pricing, and expect AI to keep moving up the ladder.
Antti Innanen-tech lawyer, legal design enthusiast, and AI geek-has also published a new book, "Prompted: How to Create and Communicate with AI" (Routledge).
Next step
If you want structured, practical training to build these workflows and skills, browse courses by role here: Complete AI Training.
Your membership also unlocks:
