Just Because Legal Can Build It, Doesn't Mean It Should: The Real Build vs. Buy Question for General Counsel
By Paula Pepin, Founder | Executive Coach | Former General Counsel, The GC Collective
A leadership perspective on AI, legal technology, and the real build-versus-buy question facing General Counsel.
The Question Every GC Is Starting to Ask
A few weeks ago, Anthropic announced Claude for Legal.
Almost immediately, I started hearing a familiar question from General Counsel:
"If AI can build software, should we be building more of our own legal technology?"
It is a reasonable question. For years, the answer was relatively straightforward. Most legal departments lacked the technical resources required to build sophisticated internal systems, so purchasing purpose-built technology was often the most practical path forward. The decision typically centered on selecting the right vendor, securing budget, and managing implementation.
Today, that equation feels different.
AI Has Changed What's Possible
A lawyer can sit down with Claude, Cursor, or another AI coding assistant and create something useful in a matter of hours. Intake forms, matter trackers, workflow automations, document generators, and knowledge tools that once required a development team suddenly appear within reach. For many legal leaders, this shift is both exciting and empowering. The possibility of solving long-standing operational challenges without waiting for engineering resources or purchasing another software platform feels like a meaningful change in what legal departments can accomplish on their own.
Yet I am not convinced the most important question is whether legal departments can build these tools. The more important question may be whether building technology is where legal teams should be focusing their time and attention.
The Hidden Cost of Building
AI has dramatically lowered the barriers to software creation, but it has not changed the reality that leadership capacity remains finite. Every hour spent designing workflows, testing automations, troubleshooting issues, and maintaining internal tools is an hour not spent advising the business, developing talent, strengthening stakeholder relationships, or shaping strategy. As legal departments continue to evolve from legal advisors into business leaders, that distinction matters.
Building Software Is Easy. Owning Software Is Hard.
Recently, Streamline AI published a thoughtful guide examining the build-versus-buy decision for AI-powered legal intake systems. The paper highlights a reality that many technology leaders understand well: creating a working prototype is only a small part of what it takes to create a sustainable business application. Security, permissions, integrations, governance, reporting, maintenance, and long-term ownership often require far more effort than the initial build itself.
Whether or not one agrees with every conclusion, the guide raises an important point. AI has reduced the effort required to create software. It has not eliminated the responsibility of owning it.
That distinction matters for legal departments.
Lawyers and Technologists Solve Different Problems
Throughout my career, I have met a small number of lawyers who were exceptionally strong operators and technologists. They understood process design, systems thinking, and how to translate business problems into scalable solutions. They were capable of looking beyond the immediate need and designing processes that could evolve alongside an organization.
Those individuals were the exception rather than the rule. Most lawyers built their careers on legal expertise, judgment, stakeholder management, and the ability to navigate risk in complex environments. While those capabilities often make exceptional business partners, they are fundamentally different from the skills required to design, maintain, and scale technology solutions over time.
Yet today, I see many legal teams investing significant amounts of time experimenting with AI-generated applications and workflows. Some of these experiments will undoubtedly create value. Others may produce something that works well enough for a demonstration but struggles under the realities of production use.
This is not because lawyers are incapable of building technology. In fact, many legal professionals are proving remarkably adept at using AI tools to create useful applications. The challenge is that building technology and operating technology are fundamentally different disciplines. Creating something that works is only the beginning. Ensuring that it remains secure, reliable, compliant, and useful as organizational needs evolve is where the more demanding work often begins.
A functioning intake form is not the same thing as an intake system. A contract generator is not the same thing as a contract lifecycle platform. A workflow that works for one lawyer is not necessarily a workflow that can support an entire legal department. The challenge becomes even more significant when legal teams begin managing highly sensitive information, privileged communications, retention obligations, audit requirements, and increasingly complex AI governance expectations.
The Real Leadership Question
There is also a broader leadership question that receives far less attention in these discussions. As General Counsel continue to take on greater responsibility as business leaders, the challenge is no longer simply deciding what can be built.
The challenge is determining: Where does the legal function create the greatest value? and then allocating resources accordingly.
Every Build Decision Is a Resource Allocation Decision
Every hour spent building internal tools is an hour not spent on strategic priorities. As General Counsel continue to take on broader business responsibilities, their greatest contribution is rarely writing software. It is influencing decisions, managing risk in the context of business objectives, building high-performing teams, and helping organizations navigate complexity. The question is not whether legal leaders can learn to build technology. The question is whether doing so represents the highest-value use of their expertise.
Innovation Doesn't Require Becoming a Software Company
Every hour spent building internal tools is an hour not spent on strategic priorities. As General Counsel continue to take on broader business responsibilities, their greatest contribution is rarely writing software. It is influencing decisions, managing risk in the context of business objectives, building high-performing teams, and helping organizations navigate complexity. The question is not whether legal leaders can learn to build technology. The question is whether doing so represents the highest-value use of their expertise.
That does not mean legal teams should avoid building altogether. In fact, I suspect the most successful departments will become much more comfortable creating lightweight solutions, testing ideas quickly, and using AI to solve targeted operational problems. AI offers tremendous opportunities to experiment, learn, and improve how legal work gets done.
But there is a meaningful difference between using AI to accelerate innovation and using AI to become a software company.
The Capability Legal Teams Need Next
The legal departments that thrive over the next decade will likely develop a new capability: understanding enough about technology to make sound build-versus-buy decisions without confusing technology experimentation with technology ownership. They will know when building creates a genuine competitive advantage, when purchasing a purpose-built solution makes more sense, and when a problem simply does not warrant the investment of time and attention required to solve it internally.
The Leadership Challenge AI Actually Presents
AI is making technology increasingly accessible. What remains difficult is deciding where to apply it. The most successful legal leaders won't necessarily be the ones who build the most tools. They will be the ones who understand where technology creates meaningful leverage for their teams and where it risks becoming a distraction from higher-value work.
Because in the end, the most important build-versus-buy decision may not be about software at all. It may be about how legal departments choose to invest their finite time, attention, and capacity. And as AI continues to expand what is possible, that question will become far more important than whether something can be built in the first place.
Not whether we can build. Whether building is the highest and best use of our attention.
