Vibe Coding Tools Give BigLaw Lawyers The Power To Build Their Own Software

Non-developer lawyers and legal operations professionals inside major law firms are building functional software applications using AI-powered coding platforms, bypassing IT departments and outside vendors entirely.

Tools including Anthropic’s Claude Code, Cursor, and GitHub Copilot have lowered the barrier to custom software development so dramatically that staff without technical backgrounds are creating working applications through natural language prompts alone.

The movement, widely referred to as “vibe coding,” represents a significant shift in how BigLaw firms approach technology, and its full implications for the industry are only beginning to emerge.

The trend gained visible momentum in late 2025 when Jamie Tso, a senior associate at Clifford Chance, began publicly sharing sophisticated legal AI tools he had built independently, replicating functionality from major legal tech vendors but customised to his firm’s specific workflows.

In a January 2026 interview with Artificial Lawyer, Tso described how his journey started after his firm enabled no-code automation tools through their existing Microsoft subscription.

Legal operations teams have positioned themselves as early adopters of this trend, with practitioners arguing it fundamentally changes how internal teams deliver value to their organisations.

Patricija “Patty” Corey, legal operations manager at cybersecurity company HUMAN, captured the shift clearly: “The teams that stand out will use AI, vibe coding, and light tech skills to build and iterate on real solutions quickly, without waiting on outside help.”

Lawyers at Debevoise & Plimpton used vibe coding to create an AI policy training game, a fully functional web application built entirely through natural language prompts, as documented on the Debevoise Data Blog in September 2025.

At Linklaters, associate James Phoenix built an AI time recording tool as a personal side project that has since been deployed across the firm, demonstrating how grassroots development can scale into enterprise-wide solutions.

Applications emerging from these efforts range from billing analytics dashboards and matter budgeting tools to conflict check enhancements that layer firm-specific business rules on top of existing systems.

Chris Bridges, co-founder of Tacit Legal and one of the creators of vibecode.law, offered a grounded assessment of what the technology can realistically achieve in a legal environment.

Bridges stated: “We’re realistic about what vibe-coding can achieve. Projects built by non-developers through AI-assisted coding aren’t production-ready out of the box. But that’s not the point.”

He added that vibe coding serves as “a catalyst for change,” allowing domain experts to demonstrate ideas rather than simply describe them, and to validate concepts before committing to full development investment.

Scott Kveton, CEO of CaseMark, launched the Thurgood platform specifically to address the gap between prototype tools built by lawyers and the infrastructure required to deploy them at scale within firms.

In a January 2026 LawSites interview, Kveton said: “You built something useful. We’ll help you get it into the hands of the people who need it. Not every attorney will build their own apps. But the ones who do shouldn’t have to become infrastructure experts to share them.”

The rise of vibe coding is forcing law firms to revisit the longstanding build versus buy question, a debate that intensified considerably in early 2026 as more homegrown tools demonstrated practical commercial viability.

For decades, law firms defaulted overwhelmingly to buying software rather than building it, operating on the assumption that legal services delivery and software development were fundamentally separate disciplines best kept apart.

That assumption is now under serious pressure, as lawyers with no formal development background produce tools that rival the functionality of established legal technology vendors at a fraction of the cost and time.