You don't need to understand how large language models work

You don't need to understand how large language models work. You need to understand three things: AI changes your firm's economics, your ethics obligations, and your competitive position. Firms using AI are 3x more likely to report revenue growth (Clio 2025), 72% of corporate clients now evaluate AI capabilities when selecting counsel (Harvard Law 2025), and using AI without governance is 'almost certainly' malpractice (ABA Opinion 512).

This is the one-page version. Everything a managing partner needs to make the decision, allocate the budget, and protect the firm -- without the technical jargon.

The Business Case: Why This Matters Now

The numbers that justify AI investment at your firm:

Revenue impact: Firms using AI report 3x higher likelihood of revenue growth (Clio 2025). AI-using firms show 25% higher realization rates. Average billable hour recovery: 1.2 additional hours/attorney/day. At a $350 blended rate, that's $109,200/attorney/year in recovered capacity.

Client demand: 72% of corporate legal departments have asked their outside counsel about AI (Harvard Law 2025). 34% of RFPs now include AI questions. 41% of GCs would consider switching firms over AI capabilities.

Competitive risk: Your competitors are adopting AI. If the firm across the street delivers research in 2 hours that takes your team 6, you're losing on speed, cost, and client satisfaction simultaneously. This gap compounds every quarter.

The investment: For a 20-attorney firm: $40,000-80,000/year for tools, plus internal implementation time. Expected return: $500,000-1,000,000/year in recovered capacity and retained client revenue.

The decision timeline: Not 'should we adopt AI?' but 'can we afford another quarter without it?' Every month of delay is a month of compounding competitive disadvantage.

The Risk Framework: What Can Go Wrong

AI risk at law firms falls into four categories. All are manageable with proper governance:

1. Malpractice risk. ABA Opinion 512: using AI without understanding its limitations is 'almost certainly' a violation of Rule 1.1. At least 8 attorneys sanctioned for unverified AI output. Mitigation: Written policy, verification requirements, training documentation.

2. Privilege risk. Consumer AI tools (ChatGPT free, Claude free) may waive privilege through voluntary disclosure. Mitigation: Enterprise tools only, signed DPAs, no client data in consumer AI.

3. Billing risk. Billing 5 hours for 1 hour of AI-assisted work violates Rule 1.5. Mitigation: Bill actual time, transparent narratives, consider value-based fees.

4. Insurance risk. Malpractice carriers are adding AI questionnaires. Firms without governance may face exclusions or premium increases. Mitigation: Written policy, documented training, written confirmation from carrier.

All four risks share the same mitigation: a written AI policy, enterprise tools, verification workflows, and documentation. This takes 30 days and $5,000-15,000 to implement. The cost of not doing it is uncapped.

The Implementation Roadmap: 90 Days

You don't need to understand the technology. You need to make three decisions and delegate the execution:

Decision 1: Approve the policy (Week 1-2). Your firm needs a written AI policy covering approved tools, data handling, verification requirements, disclosure, and billing. Have your risk partner draft it using ABA Opinion 512 as a framework. Review it. Approve it. Distribute it.

Decision 2: Select and fund the tool (Week 2-4). Start with one tool. For most firms, that's Lexis+ AI (if you're a Lexis firm) or CoCounsel (if you're a Westlaw firm). You're probably already paying for access. If not, budget $150-250/user/month. Fund a 5-8 person pilot for 30 days.

Decision 3: Mandate training (Week 4-8). Require all attorneys to complete 2-hour AI awareness training (Tier 1). Require pilot users to complete 6-8 hours of hands-on training (Tier 2). Document completion. This is your Rule 5.1/5.3 supervisory obligation -- you can't skip it.

After 90 days: Review pilot data. Make the business case with real numbers from your own firm. Roll out by practice group. Assign ongoing governance to an AI committee or designated partner.

Your role as managing partner: Approve the policy, fund the tools, mandate training, and hold people accountable. You don't need to use AI yourself (though you should understand it). You need to ensure your firm uses it responsibly.

The Economics Shift: What Changes for Partners

This is the part most managing partners resist, because it touches compensation:

The billable hour problem: AI makes lawyers faster, which means fewer billable hours per task. If you don't adjust your pricing model, AI reduces revenue on every task it touches. The solution isn't avoiding AI -- it's adapting pricing.

The pricing evolution: Fixed fees for AI-assisted routine work (research memos, first-draft contracts, standard reviews). Hourly billing for complex work (trials, negotiations, novel legal questions). Hybrid models that capture AI efficiency as margin. Firms using AI on fixed-fee matters report 40-60% higher effective hourly rates.

The leverage shift: You'll need 20-30% fewer junior associates over 5 years, but the ones you hire will do higher-value work at higher rates. Associate-to-partner ratios decrease. Per-attorney revenue can increase if you handle more matters per attorney.

The compensation implication: Partner compensation tied exclusively to billable hours penalizes efficiency. Add metrics for: matters handled, client satisfaction, effective hourly rate (revenue/actual hours), and practice development. Partners who handle 30 matters efficiently should earn as much as those who grind 2,200 hours on 15 matters.

The bottom line for partners: AI doesn't threaten partner income. It threatens the *model* that determines partner income. Adapt the model and you come out ahead.

The One-Page Decision Summary

Why: Clients expect it (72% asking). Competitors have it. Revenue impact is 3x. Cost of inaction exceeds cost of action.

What it costs: $40,000-80,000/year for a 20-attorney firm. $200,000-500,000/year for a 100-attorney firm. ROI: 3-5x within 12 months.

What can go wrong: Malpractice (unverified output), privilege waiver (consumer tools), billing complaints (inflated hours), insurance gaps. All mitigated by policy + enterprise tools + verification + training.

What you need to do: 1. Approve a written AI policy (Week 1-2) 2. Fund enterprise AI tool licenses (Week 2-4) 3. Mandate attorney training (Week 4-8) 4. Run a pilot and measure results (Week 4-8) 5. Roll out by practice group (Week 8-12) 6. Assign ongoing governance (Month 3+) 7. Audit annually

What you don't need to do: Understand the technology. Become an AI expert. Overhaul your entire firm. You need to make the decision, fund it, and hold people accountable.

The one question to ask yourself: If a competitor firm presents to your biggest client next month with a clear AI capability story and you can't match it, what happens? That's your answer.

The Bottom Line: Three things managing partners need to know: AI changes your economics (3x revenue growth for adopters), creates new ethics obligations (ABA Opinion 512 makes governance mandatory), and shifts competitive position (72% of clients evaluating AI capabilities). Implementation takes 90 days and $40,000-80,000/year for a 20-attorney firm. The cost of waiting exceeds the cost of action every quarter you delay.