Johnson Stokes & Master integrates Microsoft 365 Copilot into legal workflows while maintaining human oversight

Law firm JSM uses Copilot agents for legal drafts, cutting prep time by 40-60%. Lawyers retain full control over client-facing work under a strict governance model.

Categorized in: AI News Legal
Published on: Jun 29, 2026
Johnson Stokes & Master integrates Microsoft 365 Copilot into legal workflows while maintaining human oversight

Johnson Stokes & Master (JSM), one of Hong Kong's oldest law firms, has built a suite of internal AI agents powered by Microsoft 365 Copilot under a governance-first model that keeps lawyers in control of all client-facing work. The deployment-developed with Microsoft and now used across the majority of enabled users-shows how a regulated legal practice can adopt AI to speed routine tasks, improve consistency, and sharpen strategic advice without sacrificing professional accountability.

Scaling expertise without sidelining judgment

Lawyers handling high-volume advisory work often spend hours on repetitive but essential steps: organizing matter context, identifying issues, and drafting first-pass advice. JSM's Employment Legal Advice Copilot Agent tackles that workload directly. It structures instructions, flags key risks, and generates a first draft that lawyers then review, refine, and finalize 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 firm recorded a 40-60% reduction in time spent preparing first-pass advice for standard employment queries, along with a 25-35% drop in substantive rework during supervisory review. On average, lawyers save one to two hours per matter, time now redirected toward legal analysis, client engagement, and strategic discussion.

Embedding AI into everyday legal work

Copilot has become part of JSM's daily rhythm well beyond the initial pilot. Lawyers use it to summarize lengthy case files, capture client instructions in Microsoft Teams with intelligent meeting recaps, and stay current on regulatory shifts through built-in research tools. The firm's focus remains on outcomes-shorter turnaround times, more consistent first drafts, and more capacity for the strategic advice that clients value most.

JSM has also developed specialist agents for legal and operational needs. A Virtual Library Assistant surfaces internal knowledge bases, precedents, and templates. A Billing Guidelines Agent reviews invoices against firm policies to flag potential compliance issues early and reduce downstream rework. These agents sit inside the Microsoft 365 environment-Outlook, Teams, Word-where lawyers already work, avoiding the disruption of standalone tools.

Governance-first design

The firm established clear guardrails from the start. Agents do not provide autonomous legal advice, do not self-learn from client data, and never bypass professional review. Lawyers remain fully accountable for every client-facing output. "For us, governance was never an afterthought," Choy added. "We were deliberate about defining what the agent could and could not do, while ensuring lawyers remained firmly in control throughout the process."

This approach has allowed JSM to innovate with confidence, aligning AI adoption with the ethical, regulatory, and professional standards that define legal practice. The firm's experience demonstrates a practical model for responsible AI use in regulated industries, one where technology amplifies human expertise rather than replacing it.

Why this matters for legal professionals

JSM's results offer a concrete example of how AI can lift the administrative burden off lawyers without compromising the judgment that clients pay for. The measured time savings-40-60% on first drafts, 25-35% less rework-translate directly into more hours for advice that moves a matter forward. For legal teams evaluating AI tools, the case underscores that governance must be baked in from day one, and that the real value lies in augmenting expertise, not automating it. As law firms increasingly explore AI for Legal applications, JSM's model shows that scaling responsibly is possible when the technology stays firmly in service of the lawyer, not the other way around.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)