Português
EnglishHow Lawyers Use AI: Is It “Ask AI” or “Delegate Tasks to AI”?
In this field, the line between AI tools and AI workflows becomes clear: the former offers isolated efficiency gains, while the latter truly enables process reengineering.
Since August 2026, several lawyers at Beijing Deheng (Hangzhou) Law Firm have gradually integrated their years of practice into the Qwen Office platform, transforming them into reusable workflows. Even users without legal expertise can complete basic consultations with system guidance, while industry peers can leverage these processes to significantly boost case-handling efficiency and consistency.
Lawyers Cai Huijuan and Xu Yan distilled over a decade of will‑drafting experience into a “Will Drafting and Practical Guide” Skill. Users need no legal background—just about five minutes of conversational interaction—to generate a complete draft will, complete with risk warnings. This innovation has also sparked public interest and discussion around topics like virtual asset inheritance and young people drafting advance directives.
The Lian Yindi lawyer team launched the “Marriage and Family AI Expert Suite,” embedding decades of casework wisdom into fact‑梳理, asset inventory, and agreement generation, comprehensively covering common family law scenarios such as divorce agreements, property division, and child custody.
Below is presented in first‑person perspective, based on Xu Yan’s sharing:
Hello, esteemed pioneers in legal tech and fellow professionals. I’m Xu Yan from Beijing Deheng (Hangzhou) Law Firm.
Cases handled by lawyers are somewhat akin to crafting fine handiwork. The same materials handed to different attorneys can yield vastly different litigation strategies and document quality. Precision stems from experience and judgment, but the tedious part lies in the numerous procedural tasks that still require manual handling one by one.
Client‑transmitted documents, WeChat chats, audio recordings, videos—each must be reviewed, timelines organized, key issues extracted, followed by statutory research, evidence analysis, and drafting. Dozens of steps interlock; if one isn’t finished, the next cannot proceed. Valuable attorney time often gets consumed by repetitive chores like searching, excerpting, organizing, and formatting.
With AI’s emergence, many began simply “asking AI”—feeding large models data and requesting complaint drafts, legal question answers, or regulatory searches.
This does improve efficiency. Yet closer observation reveals that sometimes more than 80% of the output needs revision. Time originally spent “writing” shifts to “editing,” leaving the workflow sequence, review responsibilities, and professional oversight largely unchanged.
01 From “Conversation” to “Task Delegation”: A Fundamental Shift in AI Usage
In practice, we’ve found that the core transformation of workflow restructuring hinges on two words: shifting from “conversation” to “task delegation.” Previously, “asking AI” amounted to fragmented, ad‑hoc assignments, leaning toward immediate Q&A. Now it resembles assigning work—like briefing an assistant—with clear instructions on case context, objectives, data, steps, and deliverable standards. The AI agent then autonomously breaks down the problem, calls upon relevant skills, and submits results.
At this point, lawyers no longer need to craft complex prompts; instead, they focus on issuing clear commands, setting standards, and accepting outcomes. For instance, for fact‑梳理, where once a lengthy prompt was required, now a simple “Help me organize the facts of this case” triggers Qwen Office’s litigation suite to produce a structured output. Pre‑trial preparation follows suit—a single command initiates evidence analysis, evaluating materials from both opposing counsel and judge perspectives.
This capability relies on Skills and expert suites.
Skills aren’t mere prompts—they define trigger conditions, required data, execution steps, verification methods, and reference templates, forming replicable professional methodologies. When multiple Skills link together under unified logic, they create an expert suite, capable of guiding lawyers through entire service flows rather than just isolated tasks.
02 Real‑World Examples: Will Skill and Deheng Marriage & Family Suite
Recently, our team launched the “Will Skill” on the Qwen Office platform. Through a simple dialogue, users input personal details and wishes, and AI generates a well‑structured, comprehensive will draft based on professional legal frameworks. While media attention highlighted this feature, its true significance lies in solidifying our team’s specialized procedures, review habits, and risk awareness into a sustainable methodology.
This marks our first step in reshaping legal service delivery via intelligent office systems, extending into pro bono legal services.
Together with Attorney Lian Yindi’s team, I developed the “Marriage and Family Law Expert Suite.” Family law involves intricate statutes, emotional complexities, and diverse document types. We encapsulated years of real‑world experience into multiple Skills, building a system that covers everything from consultation to agreement drafting:
Fact‑梳理 Skill: Automatically parses communication records and attachments, extracting key timelines and points of contention.
Asset Inventory and Analysis Skill: Systematically organizes marital assets and liabilities, offering partition recommendations.
Agreement Generation Skill: Generates draft agreements tailored to different stages of the process.
This suite transforms tacit experience into systematic assets, ensuring each new case builds upon accumulated wisdom.
03 Scheduled Tasks and Web Page Publishing: Often Overlooked Yet Highly Valuable Scenarios
Beyond case management, there are two seemingly minor yet highly practical use cases.
The first is scheduled industry news updates. As specialists in capital markets dispute resolution, we must stay current. Manual research proved unsustainable over time. Now, we set automated tasks: Qwen Office weekly scrapes, aggregates, and formats beautiful legal information pages, complete with charts, source links, and table of contents. Once configured, it takes only a minute.
The second is client relationship maintenance. We’ve built a legal information web page with a backend database, regularly updating regulations and policies relevant to clients’ businesses and sending weekly push notifications. Clients simply click a link to access the latest content.
A client once remarked, “Attorney Xu, you truly care about our company’s growth.” To clients, consistent engagement feels like high‑quality service; for us, Qwen Office handles data collection and formatting, freeing us to focus on curating only the most pertinent information.
04 Turning Experience into Knowledge Assets: Every Lawyer Can Build Their Own Skill
Many lawyers perceive Skill creation as technically daunting, but in reality, we all deploy internalized “invisible Skills” daily: assessing incoming materials, determining factual completeness, evaluating evidence from various angles, identifying missing pieces, and checking deliverables before submission.
In the past, lawyer development depended on years of accumulating cases—valuable experience, yet prone to loss when personnel leave. Skills offer a new path: transforming individual experience from a disposable resource into cumulative, shareable knowledge assets, serving as a springboard for team-wide advancement.
And implementing all this on the Qwen Office platform is remarkably simple—perhaps just typing a few words launches your own professional workflow.
Conclusion
AI can serve as a search tool, assistant, collaborator, even a co‑author, but the most crucial factor remains how we choose to view AI: by harnessing technology, we can genuinely reclaim time for reflection and reserve expertise for human minds.
The First Half of 2026 Shanghai Legal Tech Achievements Release and AI‑Driven Legal Services Application Symposium was hosted by the Shanghai Bar Association’s Informationization Committee and Legal Technology Committee, organized by Alibaba’s Qwen Office, and supported by Deli Technology.
DomTech is DingTalk’s official designated service provider in Macau, dedicated to providing comprehensive DingTalk solutions. If you’d like to learn more about DingTalk applications, please contact our online customer service or reach out by phone at +852 95970612 or email at [email protected]. Our skilled development and operations teams bring extensive market experience to deliver professional DingTalk solutions and services!