Jamie Tso Interview: Vibe-Coding Your Own Legal AI Tools

2026-01-05 · Source: Artificial Lawyer · Field: Legal & Regulatory — Legal Technology (LegalTech), Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, medium

Summary

Clifford Chance senior associate Jamie Tso has developed a suite of sophisticated legal AI tools, often referred to as "vibe-coded," that rival offerings from major legal tech companies. Tso, based in Hong Kong, began his journey by attempting to automate the mapping of retail fund prospectuses against regulatory requirements using small language models, TensorFlow, PyTorch, and BERT. His efforts gained significant traction when Clifford Chance provided access to no-code automation tools like Copilot Studio and Power Automate, enabling him to build internal document-processing tools that were widely adopted. Tso has since transitioned to teaching himself to code, open-sourcing several repositories that have been forked over 100 times, demonstrating their practical value. He believes the economics of legal tech are shifting, making internal tool development more viable due to improving LLMs for coding and the rise of disposable, workflow-specific software.

Key takeaway

For CTOs or Directors of AI/ML evaluating legal tech investments, consider empowering your legal professionals with no-code and low-code platforms to develop specialized internal tools. This approach can reduce reliance on expensive, overlapping commercial subscriptions and allow for greater customization and auditability, especially for niche workflows where vendor solutions are less incentivized. Your firm could benefit from reduced software costs and increased agility in adapting to specific practice group needs.

Key insights

Internal "vibe-coded" legal AI tools can be a viable alternative to commercial platforms for law firms.

Principles

• Focus on solving real problems.
• Open-source tools foster innovation and adaptation.
• Economics of internal tool building are shifting.

Method

Start with no-code tools, then transition to coding for deeper problems. Share learnings and open-source solutions to gather feedback and foster collaboration.

In practice

• Experiment with Copilot Studio or Power Automate.
• Explore open-source legal AI tools on GitHub.
• Consider contract simulation for risk modeling.

Topics

• Legal AI Tools
• No-code Development
• Large Language Models
• Open-Source Software
• Contract Simulation

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Legal Professional, Software Engineer, AI Product Manager

Related on AIssential

• OpenAI Plans ‘Codex For Legal’ — Major tech companies are increasingly developing specialized legal AI offerings, intensifying competition in the legal tech sector.
• Wrap: Legal Spend Shock, Billions Battle, TR Deal, Legal Innovators California + More — Generative AI is rapidly transforming the legal sector, driving both innovation and client expectations for efficiency and cost savings.
• OpenAI says ChatGPT is not a lawyer, asks court to toss insurer's lawsuit - Reuters — Generative AI platforms are not considered legal practitioners and do not provide legal advice.
• Wordsmith lands $70m Series B to turbocharge legal work with AI - Sifted — Wordsmith's $70m Series B highlights a growing legal AI market focused on automating organizational legal requests for in-house teams.
• Legalweek 2026: When a profession built on prudence begins to wonder whether caution itself could become negligent, you’re no longer talking about tools. — Legal AI adoption is hindered by cultural resistance, fear, and misaligned incentives, not technical limitations.
• A new worst coder has entered the chat: vibe coding without code knowledge — Vibe coding tools enable rapid app creation but often produce insecure, unmaintainable code without developer oversight.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Lawyer.