Legal industry leaders in China explore compliance, ethics and trust risks in AI era

Simon Ye’s word was “shadow AI,” noting that if an employer does not provide safe, approved AI tools, employees will use their own, which in turn creates hidden risks around data privacy, security and compliance. For Steven Zhou, “reshaping” was the word. Zhou explained AI has fundamentally reshaped the legal industry while driving a transformation of traditional business models. The phrase that came to mind for Brian Tang was “in-house.” In Tang’s estimation, in-house lawyers are actually further along with AI adoption when compared to their law firm counterparts. Meanwhile, the word for Huandong Gao was “interesting,” as he shared his experiences of how different organizations he has worked for have taken varying approaches towards AI innovation and utilization, and thus find themselves in different places on their AI adoption journeys.

For Robert Lewis, the word that came to mind was “optimistic.” He observed that agentic AI may provide significant revenue-generating opportunities for law firms in the here and now, rather than at some distant date in the future.

AI adoption in China’s legal market is already mainstream. In China’s legal market, AI is no longer a future concept, it is already part of day-to-day legal work, according to Christie Wang. However, while the AI adoption rates are high, considering adoption alone can be deceptive. In her view, maturity is not just about how many people are using AI, it’s about how—and how well—they use it. Wang expressed concerns that only one-third of organizations surveyed feel very prepared to handle the risks, regulations, security concerns or ethical issues associated with AI. Wang expects that next phase of maturity will see AI adoption evolving towards legal practitioners using AI tools with increased skill and confidence.

Greater law firm efficiency. Steven Zhou told the panel that his Hui Ye law firm was already seeing an impact from AI as it builds its own AI-based management system. The firm, which employs more than 2,500 people across 43 offices, already is seeing greater efficiency, Zhou said.

Zhou also observed that knowledge management is very important for his firm, and traditionally it has been quite labor intensive. However, by utilizing AI tools in the organization and retrieval of legal documents and legal opinions, the firm has been spared a massive amount of manual work, Zhou said. Moreover, on the practice side, AI has helped measurably in many aspects. For instance, Zhou noted that when a firm team is handling litigation cases with complex facts, conducting due diligence, or engaging in an anti-fraud investigation project, AI can help compile facts and generate reports or litigation documents in minutes compared to many hours over a course of days, as was the case in a pre-AI world.

"Given the efficiency in both our internal operations and our service delivery, we definitely believe that a further investment in AI is cost efficient,” Zhou concluded.

The need for caution and having a human in the loop. While he is supportive of a cross-disciplinary approach and having IT professionals support in-house teams with high-tech innovations for legal problem solving, Huandong Gao still puts a premium on legal experience and expertise in the AI age. It has been his experience that if AI is relied upon without appropriate human intervention, big problems can emerge, as AI itself often cannot adequately identify the fundamental legal, compliance and liability issues at hand. Nonetheless, Gao is in favor of introducing innovative AI technology to his current company, where it has been shunned until recently, but he wants to move forward in a cautious and deliberate manner. He thinks the best way to accomplish this is to involve the most experienced and senior lawyers in the development of company know-how.

Building trust into products. While more than 90 percent of survey respondents indicated they were using AI in their legal undertakings, 39 percent of the interviewees voiced concerns about ethics and having adequate training. In Simon Ye’s view, AI systems should be addressing these concerns head-on.

First, according to Ye, a platform user should not have to figure out whether an AI system is safe or compliant. Rather, those elements need to be embedded into the system itself. That means making clear how answers are generated, where the information comes from, and what the system can and cannot do. Second, appropriate and assessable training is a best practice. If a tool requires a lot of training it is probably too complex, Ye observes. AI should feel like a natural extension of legal work, not a new skill to learn. Additionally, instead of training users to adapt to the AI systems, the AI platform should feed legal workflows, research, drafting and review processes so that the adoption becomes intuitive.

Lastly, AI systems should be guided rather than open-ended. That means products should have suggested prompts, structured output and workflows based on interfaces, so a user can adopt the AI tool safely without needing deep technical knowledge. These features would lead to more trust in AI from lawyers, according to Ye.

A window on legal training in the AI age. Brian Tang, the director at Hong Kong University’s LITE lab, pointed to some fascinating developments in the area of training tomorrow’s legal professionals who embrace AI natively. Tang described programs where university students are building AI-driven law firms, both front and back-end, with automated client intake, case management systems, as well as legal, regulatory and policy research.

The students work with law firms and large multinational in-house groups on directed AI projects for proof of concepts. So far, LITE lab students have worked with about 130 startups and are using AI for their research and coordinating their projects, according to Tang. Additionally, students are working on AI cutting-edge governance issues, with input from industry collaborators, as they grapple with some of most challenging dilemmas facing the legal community today.

A word about methodology.

The survey, which provides current and in-depth perspective on law firms and corporate legal departments, reflects insights from 810 legal professionals across the U.S., China, and nine European countries — Germany, the Netherlands, the United Kingdom, Belgium, France, Italy, Spain, Poland and Hungary.

You can access a recording of the Competing for Trust in an AI-Driven World webinar by clicking here, as well as the 2026 Wolters Kluwer Future Ready Lawyer Survey — Building Confidence in an AI era by clicking here.