Banner artwork by Shutterstock Gen AI / Shutterstock.com
Cheat Sheet
- Build a personal operating system. AI can help motivated lawyers expand into generalist coverage faster, but only if they build a personal operating system: assess, draft a first pass, escalate as appropriate, and learn.
- Create reusable legal assets. The system depends on a clear assessment process: what you will cover, what you will escalate, and what you will not touch solo.
- Treat AI like a junior teammate. You must provide context, guardrails, and verification steps.
- Expand responsibility with string guardrails. Convert one-off work into reusable assets such as playbooks, prompts, checklists, and templates, while protecting trust through confidentiality, privilege, and a visible lawyer-in-the-loop habit.
Here is a repeatable playbook to expand from specialist to generalist safely: Use AI to move faster, but keep human judgment, escalation discipline, and your own voice at the center.
I. AI changes the economics of the in-house legal team
A familiar in-house stereotype is that the solo general counsel at a small company does everything, while large-company legal teams are made up of specialists with narrow swim lanes. Larger companies can afford specialization. Smaller teams often cannot. In practice, most in-house lawyers live somewhere in between: deep expertise in a handful of areas, with the ability to address a broader range of issues.
AI will change that balance. It can make reviews, drafting, issue spotting, and analysis faster. It can also create confident errors, hallucinated legal rules, and cookie-cutter responses that miss the nuance of the business situation. It may mean that more in-house lawyers can become practical generalists. It may also mean that large legal departments can create hyper-specialized teams with enormous capacity to turn out focused work product.
For in-house counsel, the better way to view this change is as an opportunity to shape work patterns and career growth. If you are a specialist with general counsel aspirations, AI can help you broaden your scope. If you lead an in-house team, AI can help you create a safer development path for team members who are ready for broader responsibility. The point is not to fake expertise. The point is to build a system that lets you take on more while knowing when to slow down, verify, and escalate.
The point is to build a system that lets you take on more while knowing when to slow down, verify, and escalate.
II. Step 1: Build your operating system
A. Develop an approach to assessment: scope, risk, and escalation
A preset assessment process is the foundation of a personal operating system. It helps you decide where AI can safely support work outside your usual area of expertise, and where the issue requires a subject matter expert, outside counsel, or a more formal review process. This approach will be familiar to anyone who has had to run an entire legal function on their own: you simply do not know everything, so you need to know how to obtain help quickly and efficiently.
One practical way to do this is to create a simple green/yellow/red coverage map.
Core coverage (green / low risk). These are areas you can handle with light review, such as contracting within an established playbook or standard internal policies. You know the area well enough to determine whether AI output is on target, and you can identify when the facts have moved outside the standard pattern.
Shared coverage (yellow / medium risk). These are matters you can drive, but only with defined escalation triggers. Examples might include employment issues that need local counsel input, privacy addendum review that depends on current regulatory guidance, or a dispute letter where the tone and facts matter. In this zone, AI may help you frame the issue and prepare a first draft, but you may still share the output, shaped with your perspective, with someone who can double-check it.
Escalate (red / high risk). These are areas you should not own without an SME or outside counsel. Examples include fast-changing sanctions or regulatory advice, bet-the-company litigation, high-stakes investigations, or matters where the factual record is incomplete and privilege needs to be carefully protected.
The benefit of being intentional is that the coverage map can become a one-page artifact. A lawyer can use it to ask for broader responsibility or more latitude to use AI in daily work. A legal leader can use it to empower team members without creating the false impression that AI has removed the need for experience.
You simply do not know everything, so you need to know how to obtain help quickly and efficiently.
B. Build reusable assets
To increase efficiency, build standard framework materials that you can reuse repeatedly. This takes advantage of one of AI's real strengths: developing generic structure quickly. Good structure does not solve every problem, but it reduces the amount of time spent reinventing the middle of the fairway.
Issue checklists. Use AI to help identify the facts you need, risks to spot, stakeholders to pull in, and common red flags. Then revise the checklist based on real matters. The value is not the first draft; the value is the accumulation of judgment over time.
Fallback clause library. AI can be useful for suggesting standard phrasing for contracts, letters, and business communications. Over time, this can become more than a list of paragraphs in a document. It can become an agent or reusable prompt set that reflects your company positions, risk appetite, and preferred tone.
Policy and procedure skeletons. Before AI, some in-house teams took shortcuts by benchmarking against public online policies. AI can now provide a strong first structure for many policies, whether or not comparable public examples are easy to find. The lawyer's role is to shape that structure to the company's operations, risk profile, control environment, and culture.
These processes will not capture every edge case. They are not meant to. They help you cover repeatable work more quickly and reliably, while also identifying the situations that are not repeatable.
The benefit of being intentional is that the coverage map can become a one-page artifact.
C. Use AI like a junior lawyer
An assessment process helps you use AI with more confidence because you start with a clearer understanding of what you are trying to accomplish and how much you are willing to rely on the output. The better analogy is not that AI is a senior lawyer or a substitute for judgment. It is closer to a junior teammate: useful, sometimes impressive, but only effective when supervised.
The prompt = assignment memo rule
Give AI the context you would give a human. Include the business background, audience, risk posture, house positions, constraints, and desired output. Ask it to identify missing facts or clarifying questions before answering when inputs are incomplete. In the low-risk zone, you will usually be able to provide better instructions and evaluate the response more quickly. In the yellow and red zones, the gaps in the prompt are often a signal that you need more facts or another reviewer.
Voice is a credibility asset; do not outsource it
Use AI to draft toward your voice, not as your voice. Business partners do not trust the legal team because every communication sounds formal and comprehensive. They trust the legal team because the advice is practical, calibrated, and recognizable. A quick spot-the-bot review can help: Is this too formal? Does it use generic legalisms? Does it fail to frame the business decision? Would I actually say this to a VP or functional leader?
One useful exercise is to have AI review your prior writing and identify the features of your style, then use those features as instructions for a first pass. But the last pass should still be yours. If your voice disappears, so does part of your credibility.
Use AI to draft toward your voice, not as your voice.
Lawyer-in-the-loop: a visible verification habit
Require citations or quotes for anything that looks like a legal rule, contractual requirement, or factual assertion. We have seen AI summarize important language when the stronger answer required direct quotes. In that situation, the human-in-the-loop role was not cosmetic; it was the step that made the communication accurate and effective.
Cross-check against primary sources and current guidance for fast-changing topics. If you have access to more than one AI platform, different models can sometimes help identify weaknesses in each other's output. But at this stage of AI, nothing substitutes for careful human review of cited references and the underlying source material. Then create a feedback loop: what AI got wrong should update your checklist, prompt, or escalation trigger.
D. Learn deliberately; build competence beyond the model
To become a trusted generalist, you cannot depend only on AI. That is the path that lets AI take over the parts of the job that should become your own judgment. To be trusted, you need experience and perspective on sensitive issues, and you need enough substantive competence to supervise AI output.
One practical method is to look across the programs an in-house legal team commonly covers and deliberately track your exposure to them. A starter list might include: contracting; compliance, security, and certifications; privacy; product counseling; IP and open source; anti-corruption and code of conduct; sustainability; AI governance; employment; marketing and corporate communications; litigation; incident response; investigations; and risk management.
For each area, track your experience and comfort level. Then choose where to invest. Participate in an ACC network or other professional community. Obtain targeted training or certification where useful. Build a support network of subject matter experts. Find best-in-class frameworks and controls, such as the U.S. Sentencing Guidelines for an anticorruption compliance program structure. The goal is not to become an expert in every area overnight. The goal is to become deliberate about moving from exposure to supervised work to independent judgment.
While doing this, remember your goals: if you are developing yourself as a generalist, you do not need doctorate-level expertise in any one area.
E. Guardrails: ethics, confidentiality, and privilege in the AI workflow
Remember key guide rails for ethics, confidentiality, and privilege. Do not paste proprietary, personal, or privileged material into unapproved tools. Be explicit about whether AI is being used for drafting, research support, or advice development, and preserve privilege protocols. Document how outputs are validated, especially for novel areas. The broader your scope becomes, the more important these habits become.
To be trusted, you need experience and perspective on sensitive issues, and you need enough substantive competence to supervise AI output.
III. Step 2: Intentionally expand your scope
A. Expand into areas that naturally connect
Once the operating system is in place, look for expansion moves that build on one another. The point is not to volunteer for everything. The point is to choose matters that are safe enough to handle responsibly, visible enough that people notice the work, and practical enough that you can reuse what you learned next time.
Volunteer for a one-off project outside your lane. This might be an incident response role, a dispute letter, an investigation support assignment, or a new product initiative. Say yes to helping structure the problem, gather facts, and prepare first-pass materials. Explicitly defer final calls that require specialized legal judgment. Afterward, create a checklist or template so the next project starts further ahead.
Expand from your contracting strength. A lawyer who knows customer contracting can move into procurement, channel, partner, or vendor work. A lawyer who knows procurement can move toward customer negotiations. Adjacent issues such as privacy, AI terms, regional laws, security reviews, and audit rights become bridges. Say yes to the adjacent issue; defer where the law or business risk is outside your coverage map; create a fallback library or prompt for next time.
Own a business unit slice or executive support lane. This gives you whole-problem exposure rather than isolated legal tasks. You see how sales, product, finance, security, compliance, and operations interact. That is where generalist judgment develops. It also gives you a place to build trust through repeated small interactions rather than a single dramatic assignment.
B. Manage the mental load
Generalist work can become all judgment, all the time. AI may reduce drafting time, but it can also increase the number of questions a lawyer can plausibly touch. That creates mental load. Here are some ideas for managing that:
- Use triage windows so every new issue does not interrupt deep work.
- Use AI for a fast first pass, but set a stopping point so you do not spend more time reviewing and refining the output than needed for the task.
- Create talking-points to manage speed while preserving trust, such as: "I can give you an initial view now, but I want to confirm one point before we treat it as advice."
These habits help the lawyer maintain sanity and remain responsive, without pretending to resolve everything instantly.
Once the operating system is in place, look for expansion moves that build on one another.
IV. Why this matters
The best generalists are not people who know everything. They are people who structure problems quickly, understand the business context, know when to escalate, and keep learning. AI can accelerate that process, but it cannot replace the judgment that makes the legal function trusted.
Good judgment and reliable process builds trust. Trust leads to more interesting work. More interesting work creates visible impact. For lawyers who want to grow from specialist to generalist, that career lift is the byproduct of the system you create.
Author’s Note: This article grew out of a collaborative effort with my colleagues on the Everbridge legal team. My sincere thanks to Jess Berry, Emily Noggle, and Brienne Carpenter for sharing their perspectives, challenging and refining the ideas, and contributing to the development of the article.
Disclaimer: The information in any resource in this website should not be construed as legal advice or as a legal opinion on specific facts, and should not be considered representing the views of its authors, its authors’ employers, its sponsors, and/or ACC. These resources are not intended as a definitive statement on the subject addressed. Rather, they are intended to serve as a tool providing practical guidance and references for the busy in-house practitioner and other readers.
