Across firms, we’re seeing a consistent pattern.
Associates generally see the value of GenAI. Yet associate GenAI adoption remains inconsistent. When they use it, they say it improves early drafts, helps organize information, and saves time.
And yet—even in environments where firm-approved tools are available, training exists, and participation is encouraged—usage remains inconsistent.
This isn’t about awareness.
It isn’t even about skepticism.
It’s about something structural.
What we’re observing
When we look at how associates engage with GenAI, a few broad patterns emerge:
- Some experiment occasionally but never turn GenAI into a daily habit.
- Some may be using tools for personal benefit outside the firm’s ecosystem, making usage difficult to see or support.
- Others arrive with GenAI exposure from law school but struggle to translate that experience into client-facing work.
- And a large group doesn’t engage at all.
The challenge isn’t any one group.
It’s that even associates who see the value struggle to integrate GenAI into how work actually gets done.
For firms expecting efficiency gains or risk reduction from GenAI, this creates a gap between investment and impact.
The underlying tension
What’s becoming clearer is that GenAI creates a unique kind of friction for associates—one that traditional rollout strategies don’t address.
Associates are:
- still building core legal skills,
- working in high-stakes, billable environments,
- and closely observing what is rewarded and what is not.
Those forces sit at the intersection of training, supervision, billing, and evaluation.
Associates cannot resolve these tensions on their own.
This is why adoption is harder than it looks—and why it doesn’t respond to simple fixes.
Four dynamics that shape associate GenAI adoption
We don’t have definitive answers yet, but a few dynamics consistently show up across firms.
1. Skill development vs. acceleration
Associates are expected to learn how to research, analyze, and draft—thoroughly, defensibly, and under scrutiny.
When a tool accelerates those tasks, it introduces an uncomfortable question:
Am I strengthening my skills—or bypassing them too early?
Several associates describe this tension using a familiar analogy: using a calculator before you’ve really learned how math works. The concern isn’t that the tool is wrong—it’s whether relying on it too soon undermines the foundation they’re still expected to build.
Without clear guidance on how GenAI supports skill development rather than shortcuts it, many associates default to familiar methods—especially when they know they may need to explain or defend their work.
2. High-stakes work changes the calculus
In law school, experimentation is low risk.
In practice, work is reviewed, billed, and tied to reputation.
When GenAI use isn’t clearly addressed in review or billing conversations, associates face practical uncertainty:
- How much reliance is appropriate?
- How do I explain my process?
- How will this be judged by the partner reviewing my work?
That uncertainty is amplified when the supervising partner is not very familiar with GenAI themselves. An associate cannot afford to introduce additional risk into already high-stakes work, so the safest choice is often not to use the tool at all.
3. Cultural signals outweigh formal messages
Associates don’t just listen to what firms say—they watch what gets rewarded.
If the clearest path to strong evaluations, good assignments, and advancement remains high billed hours, GenAI creates a dilemma:
use it and finish faster—or work in ways the firm already knows how to evaluate.
Unless associates are working for partners who actively use and promote GenAI themselves, explicit signals that efficiency and higher-quality GenAI-supported work are valued alongside hours are often missing.
This isn’t resistance. It’s risk management.
4. There’s no visible progression
Many associates aren’t sure:
- where GenAI is appropriate early in their career,
- what “good use” looks like in practice,
- or when it shifts from a learning aid to a legitimate work tool.
When that progression isn’t visible, GenAI remains optional rather than expected.
And optional tools rarely become habits—especially in high-pressure environments.
Without new habits, firms don’t see meaningful operational change.
Why common fixes fall short
- More training increases knowledge, not confidence.
- Mandatory usage creates compliance, not judgment.
- Broad encouragement without changes to evaluation or incentives often backfires.
None of these address the structural tensions associates are navigating between learning, billing, supervision, and advancement.
What’s showing promise (in small but meaningful ways)
A few approaches appear to reduce friction:
- Peer normalization
Pairing associates with slightly more experienced colleagues who use GenAI regularly—but imperfectly—helps normalize use without pressure. - Narrow, low-risk entry points
Clear guidance like “use GenAI only to summarize long email threads” reduces ambiguity and builds confidence through repetition. - Explicit leadership signals
When senior lawyers talk openly about how they use GenAI, how they review GenAI-supported work, and how efficiency is evaluated, uncertainty drops.
These interventions don’t force adoption—they make it safer.
What we’re still learning
The gap in adoption isn’t convincing associates that GenAI is useful.
The real challenge is helping them move from:
“I see the value”
to
“I use this confidently as part of my work.”
That transition has less to do with training and much more to do with:
- clarity about expectations,
- visible role models,
- aligned incentives,
- and consistent signals across the firm.
If you’re working on associate GenAI adoption, we’re curious:
- Where do associates hesitate?
- What uncertainties keep resurfacing?
- What changes have actually shifted behavior?
Because this is a challenging problem—and it’s not one firms will solve through tools alone.
Associate GenAI adoption is not a rollout problem.
It’s a coordination problem—between learning, supervision, billing, and evaluation.
Until those signals align, even well-intended efforts will struggle to produce consistent results.