CBLJ FORUM SHANGHAI 2025

CHEN YI shared with the audience the error tolerance mechanism that was introduced in the set of guidelines that the State Council published to promote the high-quality development of government investment funds.

He spoke about the scope of the error tolerance mechanism, saying that this has paved the way for state-owned enterprises to establish their own error tolerance frameworks, embedding the concept into their bylaws, fund structures and investment strategies.

Jiang Wei spoke about his experience of conducting due diligence when a fund of funds (FoFs) invests in its sub-funds, and talked about the key issues that need to be considered for different types of sub-funds. Jiang says the FoF manager dictates the selection and due diligence of its sub-funds, and they must “keep their eyes wide open” when selecting better-performing funds for its investors.

Drawing on her own practical experience, Li Huiling shared insights on the exit process for FoFs. Exiting is always the end goal for financial investors, and there are two main exit strategies: passive and active.

A passive exit means that the sub-fund has direct control of the project, and the FoF will step in when the sub-fund formulates an exit strategy. An active exit means exiting through the transfer of secondary fund interests.

Betty Xiang shared strategies for investing insurance capital, a kind of patient capital. In recent years, Ping An has gradually shifted towards an M&A-driven investment strategy. M&A transactions are mainly completed through co-operating with other investment institutions or partnering with industrial capital.

Chen Yimin analysed the evolving trends of FoF and M&A funds in recent years, from the perspective of a legal practitioner. Before 2023, he said, there were more blind-pool funds and direct investment funds. After 2024, there has been an increase in alternative funds. At present, FoFs are active and the scale of M&A funds has increased compared with previous years. There would be a lot of potential in these two markets in the next two years, Chen said.

Photos

HOW TO DEAL with non-performing assets has long been a major challenge that companies must tackle.

On this issue, Qian Xin shared his experience in resolving difficulties in enforcement, the obstacles encountered in handling non-performing debts, and the approaches adopted in the past two years when confronted with such challenges. He suggests that new responsible parties be brought in to increase the likelihood of enforcement and to use mediation institutions’ services.

Zhu Xingdong spoke about the business practices at WeBank on dealing with non-performing loans. These include the challenges in settling cases involving the disposal of micro and small non-performing assets in local courts, as well as using a new notorisation and enforcement model, which allows debtors to first receive interest reductions, and obtain instalments to resolve their debts.

Candy Lu spoke about collecting and recovering assets, using her experience at Porsche Financial Services as an example. She says risk prevention is far more important than risk response, and encouraged in-house counsel to be involved in advance. They should participate in formulating the procedure of various stages, from asset management to credit assessment.

Liu Chenqu talked about resolving difficulties in the asset management sector. He says the industry is currently facing the dilemma of “winning the judgment but losing at the enforcement stage”. To tackle this issue, he suggested that asset managers could approach asset disposal from an investment banking perspective.

Cui Lisha shared her experience in resolving enforcement difficulties in the construction sector. She says achieving a resolution through a non-contentious approach is effective, and communicating and negotiating with key stakeholders should be the first step before deciding to pursue litigation.

As for the contentious approach, she suggests pursuing litigation in multiple places to increase the likelihood of successful enforcement.

Photos

IN TODAY’S COMPLEX business environment, M&As face layers of uncertainty, making risk control the deciding factor in whether a deal succeeds or fails. Experts outline systematic strategies that span the entire life cycle of an M&A transaction.

Lilian Li believes that any assessment of an acquisition’s feasibility should start from two guiding principles: practicality and security. She stresses the need for a systematic approach to risk mapping, beginning with a clear understanding of the seller’s intentions and background. Equal attention, she says, should be given to the integrity and market reputation of the ultimate controller. Companies must also identify any special procedural requirements for particular entities and engage professional external due diligence advisers. Internally, a well-defined investment decision-making framework helps ensure that M&A projects advance efficiently, while keeping risks firmly under control.

From the perspective of outbound acquisitions, Molly Hou emphasises that effective risk control starts with transaction structure and contractual design. The framework, she says, should take into account the target country’s approval requirements, tax policies and legal protections. In terms of contracting arrangements, buyers can set up a local limited liability company as the acquisition vehicle, creating physical separation and shielding existing businesses and shareholders from potential litigation or tax exposure.

Zhang Jiangbo highlights the evolution of the legal department’s role in corporate acquisitions. In cross-border joint-venture projects, especially in the fast-growing new energy and lithium battery sectors, lawyers are shifting from passive support to active leadership. Inhouse counsel are now deeply involved from confidentiality agreements through to closing, managing legal risks while also engaging closely with business strategy to serve as a key nexus for project execution.

Jasmine Chen echoes this view, adding that “completion of a deal marks only the beginning of long-term integration; in-house counsel must remain vigilant to post-merger risks”. After closing, she advises, companies should promptly review key contracts of the acquired entity to identify any abnormal or high-risk arrangements entered into before the sale, especially those designed to inflate valuation, and guard against potential fraud. It is equally vital to form dedicated integration teams early to align management cultures and systems. During any earnout or performance guarantee period, she adds, businesses should work on internalising their core competitiveness rather than depending on the seller’s short-term co-operation.

Adding a final note of caution, An Ning addresses the risk of M&A disputes escalating into criminal cases, including allegations of contractual fraud. Drawing on real-world examples, he identifies four key areas for prevention. First, ensure the genuine intent of the transaction, especially for acquisitions based on future earnings. Second, maintain independence in valuation to prevent procedural formalism or manipulation. Third, calibrate the buyer’s degree of involvement, as excessive control could weaken its legal position later. Finally, strengthen internal personnel management to prevent individual misconduct that might compromise the company’s interests. Throughout the process, he concludes, all parties should remain alert to potential criminal exposure.

Photos

AS CROSS-BORDER OPERATIONS become increasingly common, corporate legal structuring, compliance management and dispute resolution have emerged as critical issues for listed companies. Panellists share insights from their respective fields.

Hang Dongxia says the complexity and diversity of disputes involving Hong Kong-listed companies stem largely from their cross-border listing structures and mixed ownership, which frequently involve state-owned, private and foreign investors. These groups often differ in their investment objectives and in their approaches to preserving and enhancing value, a divergence that easily leads to misaligned contractual arrangements and inherent tension in legal drafting. “The greatest uncertainty lies in the validity of the agreements themselves,” she says. Hang recommends that companies start from commercial substance when structuring deals: by carefully selecting governing law and dispute resolution venues, they can at least balance the interests of all parties, even if conflicts cannot be eliminated.

Gary Wang adds that for dual-listed A+H share companies, compliance now faces multiple layers of pressure, from dual regulatory regimes to increasingly complex data compliance requirements, and above all, the tightening demands of ESG. He argues that companies must build dedicated global ESG teams and integrate sustainability requirements into long-term strategic planning. While this raises both costs and operational difficulty, such capabilities, he says, have become essential for firms aspiring to internationalise.

Joe Zhou highlights the differences in regulatory philosophy. Hong Kong’s capital market, he explains, is built on disclosure-based, ex-post supervision, where self-discipline and post-event accountability are paramount and market integrity carries special weight. “If your credibility falters, the market will abandon you,” he warns. This calls for listed companies to place continuous compliance at the top of their agenda.

When risks harden into actual disputes, matters become even more complex. Wang Wenjun points out that, in securities compliance and litigation, a single violation often triggers three-dimensional liability – administrative, civil and criminal – creating a multilayered enforcement framework. She contrasts the claims systems across jurisdictions: the Chinese mainland offers more diversified mechanisms and large-scale investor actions, whereas in Hong Kong, compensation typically occurs through injunctions pursued by the Securities and Futures Commission, resulting in a concentration of major cases. Wang highlights an important development: the Beijing Financial Court has ruled that it has jurisdiction over H-share companies if their conduct harms mainland investors’ interests, applying the new Securities Law. This marks an expansion of cross-border jurisdiction. Companies must keep a close eye on regulatory trends in both jurisdictions and strengthen their compliance and crisis management capabilities.

Yan Jun turns to the biopharmaceutical sector, noting the traits of business development (licensing-in/licensing-out) transactions, where the parties’ interests are tightly intertwined. “Owning the patent alone isn’t enough,” he explains, “You still need the original applicant’s access to existing clinical channels to complete the trials.” He cautions that overly stringent contracts can backfire: true risk sharing and aligned incentives are far more sustainable. Clauses that excessively favour one party may trigger passive resistance, such as minor technical breaches, from the other. He also stresses the importance of defining key terms precisely, such as the exact criteria for “Phase III success” in clinical trials, and suggests adding non-poaching provisions to non-disclosure agreements to create deterrence and protect the stability of long-term -partnerships.

Photos

Zhang Xuchun kicks off the session by asking how the roll-out of trustworthy data space (TDS) could address trust issues in healthcare data scenarios.

Luo Zhen’s attention centres on the two-fold meaning of a TDS concerning both the formulation of standards and their technical execution. In terms of setting national standards, “trust” is fundamentally about “unity of knowledge and action”, which guarantees that the practical operation of data exchange adheres to mutually agreed rules. Technically, paths like data encryption and privacy computing can ensure that data is usable but not visible, facilitating secure transfer. The developmental trend involves using standards to formally merge compliance mandates with technical measures. By following these standards, compliance risks can be mitigated automatically, fostering a genuinely reliable data flow environment. Luo also highlights the notion of “data trusteeship”, explaining that technology can enforce the requirement for authorised access to hosted data, enabling effective governance. This approach also serves as a lower-cost entry point for businesses into the TDS ecosystem.

Zhang Lihong notes several obstacles to data flow within the healthcare sector. Among them are enterprises’ vague data categorisation and compliance limits, which make employees hesitant to utilise data. In her view, the TDS framework and its regulations aid in verifying the legitimate sourcing of clinical data. She also looks forward to the introduction of global standards for cross-border data transfer, which is expected to offer businesses a defined route for secure international expansion within regulatory bounds. Zhang additionally raises issues concerning profit-sharing among parties in the data circulation chain, data accessibility, and the practice of data sharing.

To address the regulatory complexities multinational corporations face when transferring data across borders, Luan Gang argues that the crux is striking a balance – complying with legal mandates while optimising the utility of corporate data assets. He points out that businesses must also weigh factors such as great power competition, national security concerns and geopolitical dynamics during this process, highlighting that disease remains the paramount challenge for medical companies. Luan says: “We should make full use of the TDS to truly unite worldwide resources. This is how we can develop and deliver improved disease solutions more rapidly, for the good of humankind.”

Ma Yanran frames the challenge of medical data’s value export from a hospital’s viewpoint: “How to enable them [hospitals] to willingly and sustainably release data value for corporate use is the essential question.” She says that a TDS, through its technical design, meets the central requirements of medical organisations, including setting up accountability mechanisms and enabling compliant channels for value release. At its core, it constructs a verifiable trust consensus across multiple parties, serving as a key pillar for the secure exchange and value utilisation of medical data. 

Jiang Jun concentrates on the application of data: “The core issue confronting medical institutions is whether data can be used and how it will be used in the future.” His focus includes technical reliability and qualification certification, institutional safeguards and ethical review, as well as commercial incentives and benefit distribution. A particular warning from Jiang to businesses is the necessity of securing the legality of data sources and the compliance of the algorithms used.

Photos

Norman Luo kicks off the session by outlining his company’s AI applications. AXA Tianping, following group directives, has established its own AI compliance policy and completed the necessary assessments. Operationally, the focus is twofold: at the front end, an AI-driven motor insurance pricing system ensures sales are both precise and efficient; at the back end, AI anti-fraud models have long been employed to monitor claims risks.

To comply with financial data localisation regulations, the firm has also deployed a standalone, localised AI tool for internal staff. “I am naturally introverted, yet AI helped me prepare this speech, allowing me to stand here and express my thoughts like an extrovert. That is a significant achievement of AI,” Luo quips.

Zeng Ni points out that for enterprises wishing to implement AI within legal and compliance functions, the key lies in standardising business processes before empowering them with technology. Taking contract review as an example, the Ping An Haoyi legal team systematically categorised all contract types and extracted key review elements to build an intelligent review platform. In terms of litigation management, her team also utilises AI tools to analyse documents such as statements of claim, extracting critical information like points of contention and timelines.

She concludes that AI empowerment has helped the team significantly improve efficiency and reduce labour costs. Although AI makes mistakes and cannot completely replace human effort, it has saved a substantial amount of time overall. However, she emphasised that AI primarily saves effort in the 80% portion of the “80/20 rule” and cannot be truly given full autonomy. “AI can bring me a great deal of optimisation and empowerment, so I have always felt that it is not a rival, but an assistant,” she says.

At the mention of AI errors, the topic naturally turns to hallucinations, a major industry challenge. Kenneth Zhou believes that strict controls need to be in place during the data input and output processes because the higher the quality of data provided to the AI, the higher the value of the content it produces. Additionally, he suggests establishing a human verification step at the company policy level, to not only review the AI-generated content itself but also check whether its data sources are authentic and reliable.

On the ethical issues behind AI, Kenneth Zhou emphasises that AI tools need to possess transparency and explainability. He suggests that enterprises establish an ethics committee, particularly in highly regulated industries such as finance and healthcare, as well as creating a whistleblowing mechanism accessible both internally and externally. He also predicts the emergence of a new profession in the future: the “ethics auditor”.

Cao Haiyang also believes that AI hallucinations are essentially a matter of data reliability. “Current AI systems are all large language models,” he explains, “their operating mechanism involves inferring what the next segment of text expresses based on previously acquired data and documents.”

He says hallucinations ultimately require professional, vertically specialised training to be corrected, noting that leading tech enterprises such as OpenAI are already making efforts in this regard. However, he also offers a line of defence for AI, posing a philosophical question: AI has hallucinations, but do humans not have them too?

Ruby Zhou, meanwhile, emphasises that the widespread application of AI tools cannot replace the role of people exercising their subjective initiative. She shares an interesting phenomenon. After LAW+ provided online courses on AI to staff at all levels of a company, she initially assumed that the course would be used primarily by the younger cohort, given that this group would be most anxious about being replaced by AI. However, the backend data revealed the exact opposite result: management led the completion rate by a massive margin.

Based on this, she proposes that the function of AI is to help everyone take the “second step” more effectively, while the more critical first step – namely, the decision to improve and break through one’s limits in the face of difficulties – still needs to be initiated by oneself.

Photos