Artificial intelligence is moving from experimentation to everyday use across the legal sector, creating opportunity and risk for law firms. The challenge is no longer whether lawyers will use AI, but how firms can introduce it in a way that supports productivity, protects clients, and maintains professional standards. In this article, I’ll share how our firm approached the rollout of Microsoft Copilot, why our law firm Copilot training became a prerequisite for access, how we measured adoption and impact, and what we’ve learned about building sustainable AI capability across the firm.
This article is for law firm leaders, L&D teams, IT teams, knowledge teams, and anyone responsible for introducing AI into a legal environment. If you are wondering how to encourage adoption without creating unnecessary risk, our experience may help.
| Highlights |
|---|
| 1. AI policies matter, but training is what turns policy into practice. |
| 2. Nobody receives a Copilot license until they complete mandatory training. |
| 3. Usage data helps us understand how training changes behavior. |
| 4. Firm-wide AI skills are becoming more important than training on a single tool. |
| 5. AI is likely to accelerate the move towards outcome-based and fixed-fee pricing. |
Who is Wiley Rein LLP?
Wiley Rein is a Washington DC based law firm best known as a regulatory, litigation, and public policy powerhouse – composed of over 250 attorneys and advisors supported by more than 250 legal professional staff members. Because of its strategic location, the firm primarily focuses on matters that intersect with the federal government, federal agencies, and international trade.
Why We Chose Microsoft Copilot AI
When we started looking seriously at AI, we realized we were behind. Our lawyers were asking for AI tools, clients were talking about AI, and the legal industry was moving quickly. At the same time, legal tech trends and stories from peers at events like ILTACON made one thing clear: even firms with strict policies were finding that people were already experimenting with AI on their own.
“The biggest risk was not that lawyers would use AI. It was that they were already using it without guidance, training, or oversight.”
That created a challenge. We needed to support innovation while protecting the firm, our clients, and our reputation.
Our answer was Microsoft Copilot.
Because we already use Microsoft across the firm, Copilot allowed us to keep data within our existing environment while giving lawyers and staff access to AI tools in a controlled way. But technology alone was not enough. Training had to come first.
We built a mandatory AI learning path through Intellek LMS. Nobody receives a Copilot license until they complete the required training. We focused as much on risks and limitations as we did on features and benefits. Today, training, governance, reporting, and ongoing support all work together as part of a broader AI adoption strategy.
How We Rolled Out Copilot Training
Our firm did not begin with the technology. We began with a business problem. We knew AI was becoming impossible to ignore. Lawyers were asking questions. Clients were asking questions. Industry conversations were dominated by AI.
What changed our thinking was the growing gap between official policy and real-world behavior.
At industry events, I repeatedly heard the same message from peers:
“People are already using AI, even when their firms haven’t rolled out an AI policy.”
That observation matched what we were seeing ourselves. If lawyers cannot access approved AI tools, many will look for alternatives. Often that means free public tools that operate outside the firm’s governance framework.
Free AI tools may offer convenience, but they can create concerns around confidentiality, data handling, accuracy, and compliance.
We therefore had two choices:
- Try to stop AI adoption completely.
- Create a controlled way for people to use AI responsibly.
We chose the second option.
At the same time, we formed an AI Task Force and introduced a formal AI policy. The policy provides guardrails for acceptable use. If someone breaches the policy, the firm has a framework in place to respond appropriately.
Policy creates protection. Training creates understanding.
You need both.
Approval Process: Copilot Training Before Access
One of the most important decisions we made was simple:
Nobody gets a Copilot license by default.
Before anyone receives access, they must complete our law firm Copilot training, a mandatory learning path delivered through Intellek LMS.
This approach solved several problems at once.
- First, it ensured every user started with the same foundation.
- Second, it demonstrated that AI adoption was being taken seriously.
- Third, it allowed us to scale training without overwhelming our internal teams.
The learning path originally consisted of five law firm Copilot training videos followed by instructor-led classes. Later, we refined the model.
The live sessions proved valuable, but they also created scheduling challenges. We eventually replaced them with a dedicated Copilot training video that covered the same material and could be accessed on demand through the LMS.
The relationship became simple:
Training completion unlocks Copilot access.
This removed ambiguity and created a clear expectation across the firm.
AI Training Must Include Risks, Not Just Features
One mistake many organizations make is focusing entirely on what AI can do. Equally important is understanding what AI should not do. Our training therefore covers both opportunities and risks. Users learn practical applications, but they also learn important limitations.
Topics include:
- Hallucinations and inaccurate outputs.
- Appropriate review and validation processes.
- Ethical considerations.
- Data handling responsibilities.
- Situations where AI should not be relied upon.
The result is a more balanced understanding of AI.
“Confident AI users are useful. Informed users are safer.”
That distinction matters in a legal environment.
Building AI Skills into the Associate Journey
We do not treat AI training as a one-time event. It’s becoming part of the wider professional development journey.
For example, new fall associates receive law firm Copilot training within their first two weeks. This ensures AI literacy becomes part of onboarding rather than an optional extra.
The relationship is clear:
Early training creates consistent habits. Consistent habits reduce risk.
As AI becomes more common across legal work, waiting months or years to introduce these concepts makes less sense. Future lawyers will need to understand AI in the same way they understand legal research tools, document management systems, and knowledge resources.
Law firm Copilot Training for Different Practice Groups
General training establishes a foundation. Practice-specific training creates relevance. After users complete their initial learning path, we ask them a simple question:
“How would you like to use it?”
The answers become the basis for future training, and this creates a more user-led model. Instead of pushing generic examples, our training team develops content around actual requirements.
- A litigation team may need different scenarios than a real estate team.
- An employment group may have different priorities than a corporate team.
Learner needs drive our technology training content and relevant training drives our tech adoption. People engage more readily when they can see direct value in their daily work.
Measuring Copilot Training Impact Through Data
Training is only valuable if it changes behavior, and that’s why measurement is so important.
We regularly review training completion reports through Intellek LMS. Administrators, HR teams, and Copilot stakeholders can also monitor progress and identify adoption trends.
We also review how Copilot is being used across the organization and how usage patterns evolve after our law firm Copilot training is completed. This creates a useful feedback loop.
Instead of relying on assumptions, we can see where users are succeeding and where additional support may be required. That visibility has become an important part of our governance strategy.
Next Stage of our firm’s AI Training
One of the most interesting findings from a recent partners’ retreat was that attorneys wanted more AI training. The challenge was to determine exactly what that training should cover.
“Technology changes quickly. Training that focuses only on one specific tool can become outdated. That has led us to think more broadly about AI capability development.”
Rather than teaching people only how a particular platform works, we want to help them develop transferable AI skills that remain valuable regardless of which tools the firm adopts in future.
The concept is straightforward:
AI skills will outlast AI product training.
This is one reason initiatives such as Intellek’s AI Readiness for Legal Professionals program are interesting to us. The focus is on practical AI literacy and foundational skills rather than a single application.
We are also exploring role-based programs and monthly learning topics driven by user demand.
We are planning an “AI Week” that will provide options for training sessions and workshops relating to different AI-related use cases and scenarios.
The long-term goal is continuous learning rather than one-off training events.
How AI is Changing the Business of Law
The technology conversation is important. The business conversation may be even more important.
We are seeing very different attitudes from clients. Some clients actively encourage AI adoption. Others remain cautious because of concerns around reliability, confidentiality, and risk. That creates a complex environment for law firms.
Another challenge is the traditional billable hour model. If AI helps lawyers complete work faster, firms must answer a difficult question:
How should that efficiency be reflected in pricing?
Clients are becoming more informed about AI capabilities. Many understand that certain tasks can now be completed more quickly than before. As a result, firms may need to rethink how value is measured and delivered. This shift may take years, but the direction of travel is becoming clearer.
My Prediction About AI in Legal
AI Will Accelerate Fixed-Fee Pricing
Looking ahead, I believe AI will accelerate the move towards more client-driven pricing models.
Large clients increasingly understand what work needs to be completed and what outcomes they expect. Instead of paying purely for time, they may increasingly pay for results.
In that environment, firms will need confidence in their processes, their technology, and their people.
- Efficiency becomes an advantage.
- Predictability becomes an advantage.
- Trust becomes an advantage.
“Firms that combine strong governance, effective training, and practical AI adoption will be better positioned to succeed. Technology alone will not create that outcome. People will.”
Law firm Copilot & AI training FAQ
Here are a few questions about AI, Copilot, and training that I’ve answered quite often:
Law Firm Copilot Training
We do not encourage reliance on AI-generated legal documents without review. Human validation remains essential because AI systems can make mistakes, including factual errors or incorrect citations.
In my view, yes. AI tools are powerful, but they also have limitations. Training helps users understand both the opportunities and the risks. It creates consistency and reduces the likelihood of inappropriate use.
Training completion reports, usage analytics, and user feedback can all help. Measuring behavior over time provides a clearer picture of whether AI training is delivering meaningful results.
No. AI is helping lawyers work differently, not eliminating the need for legal expertise. Professional judgement, client relationships, strategic thinking, and ethical responsibility remain essential.
Not necessarily. We chose a training-first model where users complete a mandatory learning path before gaining access. This ensured a minimum level of competence and understanding.
For many law firms, the next step is broader AI capability development. That includes prompt writing, critical evaluation of outputs, responsible AI usage, workflow design, and understanding where specialist legal AI tools fit into the wider technology landscape.
In our experience, lawyers increasingly want practical guidance that relates directly to their role. AI training tied to real-world legal workflows often generates stronger engagement than generic demonstrations.
Paul Buckin uniquely bridges the gap between complex legal practice and modern technology infrastructure, bringing over 30 years of L&D expertise to his role in shaping the firm’s strategic technology operations. A recognized industry voice in LegalTech strategy and AI governance, he is a frequent panelist and speaker at ILTACON, where he advocates that modern legal IT leaders must evolve past standard troubleshooting to act as “strategic enablers” who integrate technology directly into law firm operations.
