|
27 August 2026 (Nashville, TN) - This fall, like every fall, large law firms will welcome new associates who are expected to acquire the writing and analytical skills necessary to become professional advisers and advocates. Associate development has long depended on both learning by doing and learning from colleagues.
Generative AI has the potential to disrupt this process by encouraging associates to outsource large portions of this work. If AI is an associate's first and last stop, rather than a tool they use to support their judgment, firms risk both creating professional responsibility issues and short-circuiting associates' learning processes.
Because of this risk, law firms must be more intentional in training their associates to be strong thinkers and writers.
New associates are particularly vulnerable to AI risks because they lack experience, face significant time pressures and want to produce polished work. Those conditions make AI's fluent answers particularly persuasive.
This is a good-news story, however. Firms can both encourage new associates to engage in independent thinking and writing and to use AI in the writing process.
To do this, they need to encourage associates to remain both the first thinker and last thinker on every legal writing project, with AI use reserved for the middle. And firms should make sure that associates' AI use is transparent, encouraged and dissected.
This "first thinker, AI, last thinker" framework recognizes that legal writing is a thinking process, not a single drafting event. It ensures that use of AI strengthens, rather than replaces, associates' thinking processes.
Associates begin with independent analysis. They then make deliberate decisions about how AI can advance their thinking and writing. Last, they conclude by critically evaluating their own work and any AI-assisted work before submitting it. AI becomes one tool in the process and a visible and transparent part of the associate's thinking and work product.
First Thinker: Preserve independent analysis before introducing AI
First-thinker work is the most important, and no technology can replace it. That means that a novice lawyer should not start with AI. The associate must be the first thinker because they are in the room with the assigning lawyer, and they are ultimately responsible for producing the work.
Thus, assigning lawyers should encourage associates to think about and internalize the assignment before engaging with AI.
After they receive an assignment, the associate should confirm their understanding of the task, perhaps by email, and seek feedback on their understanding. The assigning lawyer should share context and background — more than they think is necessary — so the associate understands how their work fits into the client matter, is encouraged to think about the work more deeply and invests in it.
If the assignment is significant, the assigning attorney should check in with the associate early on to share feedback on the associate's thinking process. At the initial meeting and any subsequent check-in, the assigning attorney should also discuss with the associate where in the assignment AI can assist.
These conversations ensure that the associate frames the legal problem both independently and in conjunction with the expert assigning attorney. If the associate goes to AI before their first-thinker work has concluded, they risk the AI answer anchoring all later thinking.
When AI anchors and becomes the basis for the associate's thinking, that limits the associate from thinking more broadly and critically about the assignment. And of course, any AI answer is limited by the prompt, background material and expertise the associate provides, increasing the chance that AI's answer will be incomplete or inaccurate.
AI in the middle: Expand the associate's thinking
Only after the associate has done the first-thinker work should firms encourage them to use AI to expand their thinking.
In this middle stage, the associate should use AI by assigning it roles: AI excels as a critic, brainstorming partner and coach. Used in this way, AI becomes a partner with the associate to advance their thinking and remain actively engaged in the analytical process. It also serves as a teacher, accelerating their learning by providing immediate feedback during the writing process.
These roles suggest the kinds of tasks firms should encourage associates to give to AI. For example, after the associate has written a draft, they may be encouraged to prompt the AI to highlight places where their writing is not clear or precise, rather than asking AI to rewrite their work.
Associates can also ask AI to provide options for how to make their writing clearer and help calibrate their voice and tone. AI can also answer questions about gaps in reasoning, or assume the role of an adversary and identify the adversary's best arguments. AI can also check grammar and fluency.
Further, this middle use is where firms should encourage associates to develop a checklist to determine whether they understand the work they're doing so they can interrogate the AI work product. Most experienced attorneys can identify fairly readily where AI misses the mark, but firms must train new associates to do the same.
Before relying on AI's work, associates should ask, "Do I understand enough about the subject to assess the AI's output, and if not, how can I verify it?" They should also ask, "Will I be able to explain and defend this analysis to my assigning lawyer?"
Firms should encourage associates to take the time they need to perform this step. They should recognize that junior lawyers' AI use is more likely to add time to a project than to reduce it.
The assigning lawyer should also discuss with the associate their middle-stage use of AI and interrogate its effectiveness. When these discussions occur, AI becomes part of the thinking process, and associates are more likely to be more deliberate and sophisticated in how they use AI.
Last Thinker: nsure the lawyer - not AI - exercises final professional judgment
The associate should always be the last thinker on a writing project. The associate must exercise independent judgment to analyze whether a final product is accurate and on point in its analysis, answers the specific questions they were asked, and meets the reader's needs. This step is where associates are most at risk of deferring to AI's answers, especially when the associate is under deadline pressure.
There are two ways firms can encourage associates to be the last thinker.
First, firms can train associates to build intentional pauses into their writing process to ensure that they are not blindly accepting the AI's output. Associates should create distance from the project so they can exercise their independent judgment before they turn it in. The associate can accomplish that distance in a variety of ways, including by the passage of time, printing out their work to encourage a tougher read or sharing their work with a close colleague.
Second, the assigning lawyer should interrogate the associate's thinking to ensure that they have internalized the learning and judgment associated with the work product. Assigning lawyers should ask new associates to explain their analysis and conclusions to ensure that they understand what they wrote and how it fits into the case or matter.
And assigning lawyers should inquire specifically about associates' AI use. They should encourage associates to record how they used AI in connection with the assignment, and the assigning lawyer should discuss with them what worked and what didn't. This helps associates learn from the experience and incorporate their learning into their writing process and subsequent work. Firms should also provide associates with opportunities to share and workshop these experiences in both small groups and larger ones.
The firms that will thrive in an AI environment will intentionally preserve the learning and apprenticeship process through which lawyers develop professional judgment and strong writing skills.
By ensuring associates remain the first thinker and the last thinker, while using AI intentionally in the middle, firms can embrace AI's efficiencies without sacrificing the judgment clients hire lawyers to provide.
| 
