Please enjoy this blog co-authored by Dr. Melissa L. Brown, Learning & Development Manager, Holland & Hart and Harriet Joubert-Vaklyes, Senior Manager of Legal Process Improvement, Michael Best.
________________________________________
Series Introduction
What Strong Leaders Carry Through AI Transformation
ILTA Leadership Development Community
Legal tech leaders sit in the rare roles where AI transformation is both a strategic mandate and a constant operational challenge. You are expected to modernize workflows, manage risk, support attorneys, and maintain legacy systems while preparing for what comes next. The technology keeps evolving, as does expectations of leadership.
This three-part series focuses on what strong leaders hold onto during that kind of transition.
Part one looks at the process discipline that keeps automation grounded.
Part two examines the judgment skills that prevent teams from leaning too heavily on system confidence.
Part three highlights the people skills that turn AI from a pilot into sustained practice.
ILTA’s Leadership Development audience does not need tool lists or hype cycles. You need leadership behaviors that work in environments where innovation has to be controlled, defensible, and aligned with how legal work actually happens.
________________________________________
Series: What Strong Leaders Carry Through AI Transformation
Part 1 of 3: Process Discipline
Title: Building Order Before Automating Anything
Author: Dr. Melissa L. Brown
Building Order Before Automating Anything
AI adoption inside law firms and legal departments rarely collapses because the technology is fragile. It collapses because the work underneath it has never been mapped, clarified, or stabilized. Leaders often hope that automation will create order where none exists. It never does – instead, it speeds up and amplifies anything that is disorganized or ill-defined.
People doing this work feel that pressure every day. They are running legacy systems while testing new ones, supporting attorneys who work in different ways, and balancing competing priorities that all claim the top of the list. The tension is real, and the underlying issues rarely start with the tools: it comes from how the work is organized.
This installment focuses on the practical groundwork that has to be in place long before a team evaluates tools. One cannot start with the abstract idea of discipline. Instead, it’s critical to recognize that the conditions that make automation dependable instead of volatile start long before you introduce technology. In other words, introducing technology (AI or otherwise) is only successful when you have a foundational understanding and documentation of the underlying process.
Why Process Maturity Dictates AI Success
AI functions best when the underlying work is steady. Legal operations often move in the opposite direction: workflows vary by practice group; knowledge lives in people instead of documentation; informal shortcuts keep matters alive during busy seasons; partners can override a process with a single email; and support teams adapt constantly just to keep production moving.
None of this is unusual. It is the way legal work survives under tight timelines, demanding clients, and shifting expectations. But these habits create weaknesses that become visible once automation enters the room. The system expects structure. Legal work depends on judgment. When those two realities collide, the system cracks first.
Leaders preparing for AI understand that the real work begins with fully understanding and then stabilizing what the technology will touch.
Setting the Groundwork for Automation
A few things need to be understood before a workflow is automated These are not theoretical requirements: they are the conditions that prevent rework, delays, and silent failure later.
1. Documentation that reflects real behavior
Many firms have plenty of documentation, yet most of it describes how work was supposed to happen at some point in the past. AI cannot operate on outdated maps. It needs a picture of the work as it exists now, including the exceptions and the messy handoffs that never made it into the official version, and/or the workarounds that have been implemented since the original mapping exercise. The aim is honest visibility.
2. Ownership that stays consistent
Automation operates cleanly when responsibility is clear and steady. If ownership drifts based on capacity, tenure, or personality, the system starts making the wrong calls and spends its time correcting for uncertainty instead of moving the work forward.
3. Standards that stay stable long enough to train the system
Teams often rely on flexible norms because the work demands it. AI needs firmer footing. Inputs, review points, and quality checks must be predictable for the system to learn. This is about giving teams a shared baseline the technology can trust.
When these conditions weaken, the system has to guess at what’s right, which can introduce uncertainty. When they strengthen, the system can be tested, measured, and relied upon.
How Legal Tech Leaders Build Order in Practice
Building order is not a documentation project. It is a reality check about where workflows break, where people compensate for gaps, and where variation introduces risk.
A useful place to start is the category of work that creates the most frustration. The tasks that constantly need cleanup. The requests that bounce between people or teams because no one is sure who owns them. The processes that fall apart whenever a key person is out of the office. These situations reveal the organization’s true operational shape, and they ultimately show exactly where AI will struggle.
Once those workflows are identified, the next step is mapping them at a workable level of detail. Too high, and the decision points disappear. Too low, and the insight gets buried. The right level is the one that exposes the moments where misalignment happens. When those points are visible, the organization can test whether automation can support the workflow or whether the process needs repair first.
The people closest to the work bring clarity here. Specialists, coordinators, and support staff understand the difference between the official story and the real one. They know how work actually moves. Their input keeps an implementation grounded and prevents drift from the practical to theory.
Before any configuration takes place, leaders should stress-test the workflow through typical scenarios, edge cases, and known failure points. These simulations expose weak seams that will crack later. If the process holds its shape during stress testing, the technology has something stable to connect to.
What Effective Leaders Reinforce During Process Stabilization
Cleaning up processes can feel like slowing down progress, especially in organizations eager to move quickly on AI. Leaders create the conditions for this work to succeed.
First, they reinforce the importance of clarity. Clarity reduces improvisation and gives teams a shared understanding of the work they are executing. That groundwork supports the technology and protects quality.
Next, they pay attention to sequencing. Legal work often involves urgent requests that skip steps. Those shortcuts create downstream instability. Leaders help teams respect the sequence so the system has a structure it can follow.
Finally, they stay close to the people doing the work. Process stabilization prevents burnout and reduces rework, and that benefit should be highlighted early and often. When leaders communicate that purpose, teams participate more fully and resist less.
What This Means for AI Adoption
Process discipline rarely gets attention because it is not dramatic or flashy work. It is slow, precise, and often invisible. But it determines whether an implementation gains traction or collapses after launch.
Organizations that give this phase the time it needs experience fewer breakdowns and cleaner data. Teams also trust the system more because the work underneath it is stable. The organizations that rush past this step spend the next several months repairing gaps they could have prevented.
AI will magnify whatever sits underneath it, and leaders shape that foundation. Once that foundation is stable, the organization can begin to explore where judgment needs to be protected as the tools enter the workflow. That is the focus of our next installment – be on the lookout!
#Leadership#BlogPost#100Level#Adoption#ArtificialIntelligence