From Legacy to AI Frontier: How Johnson Stokes & Master Built a Governance-First Copilot Model for Regulated Legal Services

As organizations across regulated industries explore AI adoption, many face the same question: how can it be scaled responsibly without undermining accountability, trust, or professional judgment?

For Johnson Stokes & Master (JSM), one of Hong Kong’s oldest and leading law firms, the answer began with a clear principle: AI should support professional judgement, not replace it.

Working closely with Microsoft, JSM has developed a growing suite of internal AI agents powered by Microsoft 365 Copilot to support lawyers’ day-to-day work within explicit governance guardrails and a clear human-in-the-loop model.

At JSM, the value of AI in regulated legal services is not simply about speed. It is about delivering more consistent work, strengthening governance, and providing faster access to relevant legal insights, while preserving clear human accountability for every client-facing outcome.

Scaling legal expertise without diluting judgement

As trusted advisors and problem solvers, lawyers need to analyze large volumes of information and deliver commercially grounded advice, often under tight timelines. Much of this work involves repetitive but critical steps, including structuring instructions, summarizing documents, identifying issues, and preparing first‑pass drafts.

AI is beginning to reshape how this work is done. Embedded within everyday workflows, it acts as an intelligent assistant, streamlining routine tasks and consolidating firm knowledge, precedents, and insights, enabling lawyers to work with greater speed, consistency, and clarity.

At JSM, this approach comes to life through a growing suite of purpose-built AI agents. One example is the Employment Legal Advice Copilot Agent, developed for Microsoft to support high‑volume employment advisory work for now, with plans to replicate the model for other corporate clients.

The agent helps organize matter context, identify key issues, and generate structured first‑pass drafts covering risks, options, and recommended next steps. Lawyers then review, refine, and finalize the advice before anything reaches a client.

“Deploying the Employment Legal Advice Copilot Agent has changed how we handle high-volume employment instructions,” said Joe Choy, Partner and Co‑Head of Employment and Benefits at JSM. “What stood out most was seeing the system perform in live matters. The output provides a genuinely strong starting point, allowing our lawyers to spend more time on the strategic judgement and client counsel that matter most.”

The impact has been significant. JSM recorded a 40–60% reduction in time spent preparing first-pass advice for standard employment queries, alongside a 25–35% reduction in substantive rework during supervisory review. On average, lawyers save one to two hours per matter, time that is redirected toward legal analysis, client engagement, and strategic discussion.

Embedding AI into everyday legal work

Today, Copilot has become part of JSM’s everyday rhythm of work.  Adoption has moved well beyond pilot stages, with Copilot now actively used across the majority of enabled users as part of lawyer-led workflows.

At JSM, adoption metrics are only one part of the picture. The firm focuses more meaningfully on outcomes—reducing turnaround times, improving the consistency of first-pass work, and enabling lawyers to spend more time on the strategic advice clients value most.

Lawyers use Copilot in different ways to support their day-to-day work. It can summarize lengthy case files and complex documents in a fraction of the time, reducing administrative workload and freeing up capacity for deeper legal analysis.

Within Microsoft Teams, intelligent meeting recaps have improved how client instructions are captured, shared, and followed up across teams, minimizing manual note-taking.

In addition, tools such as Microsoft’s built-in Researcher agent help lawyers stay up to date with regulatory developments and surface relevant insights efficiently.

Building on these capabilities, JSM has also developed a growing portfolio of specialist AI agents tailored to legal and operational needs:

• Virtual Library Assistant, connecting lawyers to internal knowledge bases, precedents, templates, and preferred research methodologies

• Billing Guidelines Agent, reviewing invoices against firm policies to flag potential compliance issues early and reduce downstream rework

Building AI with governance at the core

Rather than introducing standalone AI tools, JSM worked with Microsoft to build AI agents directly within the Microsoft 365 environment, including Outlook, Teams, and Word, where lawyers use every day.

From the beginning, the firm established clear guardrails. The agents do not provide autonomous legal advice, do not self-learn from client data, and do not bypass professional review. Lawyers remain fully accountable for all client-facing work.

“For us, governance was never an afterthought,” added Choy. “We were deliberate about defining what the agent could and could not do, while ensuring lawyers remained firmly in control throughout the process.”

This governance‑first approach has allowed JSM to innovate with confidence, aligning AI adoption with the ethical, regulatory, and professional standards that define legal practice.

A blueprint for responsible AI in regulated industries

By embedding AI directly into existing Microsoft 365 workflows, JSM has avoided the disruption and training barriers that often come with introducing entirely new systems. Instead, adoption has been gradual, practical, and closely aligned with existing ways of working.

“Microsoft continues to evolve rapidly in response to real-world business needs, and we are learning alongside that pace of innovation,” said Jonathan Voo, Senior Innovation Manager at JSM.“ For us, adopting Microsoft 365 Copilot is not about technology for technology’s sake. It is about building a more scalable and effective way to support our lawyers and serve our clients, while maintaining the professional standards and judgement that define legal practice.”

JSM’s experience demonstrates how organizations in highly regulated industries can use AI to improve efficiency, consistency, and knowledge sharing while maintaining trust, accountability and professional standards.

As AI becomes increasingly embedded in professional work, the firm’s approach offers a practical blueprint for responsible adoption: one where technology amplifies human expertise rather than replacing it.