CHINA Trends and Developments Contributed by: Ting Liu, Fang Pan and Xiangru Chen, King & Wood Mallesons
• The regulatory demand for more precise, time- sensitive evidence collection – AI delivers an unmatched advantage in speed and precision. In an environment, such as HR internal investiga - tion, where timelines are tight and data volumes are overwhelming, manual review processes are inherently slow and vulnerable to oversight. AI transforms this dynamic by enabling rapid, intel - ligent triage of massive datasets, even process - ing in hours what would traditionally take weeks. Furthermore, powered by machine learning, these AI systems continuously refine their detection mod - els based on past investigation findings, ensuring accuracy and relevance improve over time. This self-reinforcing cycle of efficiency and evolving precision cements AI as a critical force multiplier in internal investigations. Unresolved challenges and areas for improvement in AI application While AI offers transformative and remarkable advan - tages to internal investigations, their application in internal investigations is not without risks and unre - solved issues. First, the problem of AI hallucinations threatens the authenticity and admissibility of investigation results. AI hallucinations generally refer to a phenomenon wherein generative AI produces false, nonsensical or fabricated content, presents it as verifiable fact, and lends it an air of plausibility despite the absence of real-world grounding. This means that generative AI and large language models can sometimes produce convincing but incorrect, misleading or entirely fabri - cated information. In an internal investigation, such hallucinations could misinterpret the tone or intent of a communication, falsely identifying connections between individuals, or generating non-existent evi - dence, etc. This can further expose employers to chal - lenges regarding the validity of their investigations, and even give rise to labour disputes and reputational damage. Second, improper use of AI may lead to breaches of confidentiality obligations and leakage of sensitive information. In practice, internal investigations may involve a large amount of sensitive data, such as cor - porate financial records, employee’s information, and
other confidential information. When employers input these sensitive data into public AI tools in the course of internal investigations to process tasks like sorting information or generating reports, the data may be retained by AI service providers as training materials. Even if the employers delete the input records after - ward, the information cannot be completely retrieved, creating irreversible leakage risks. Third, the powerful data-processing capability of AI raises serious concerns regarding personal informa - tion protection. Internal investigations often require analysing employee-related data, such as work emails, chat records and access logs. This data all falls under the scope of personal information protect - ed by laws like the Personal Information Protection Law of the People’s Republic of China. When AI tools process such data on a large scale, they may inad - vertently overstep necessary boundaries: for instance, collecting employees’ private chat content unrelated to the investigation or analysing the personal infor - mation of individuals not involved in the case. Such practices could violate the legitimacy, necessity and proportionality principles governing personal informa - tion processing. Summary In light of these risks, while AI enhances investigative efficiency, it also introduces new challenges concern - ing accuracy, security and compliance. Employers must not treat AI as an autonomous decision-maker. Instead, it is recommended to develop a balanced, governed and accountable process for AI-assisted investigations. This process must ensure that the pur - suit of technological efficiency does not compromise fundamental principles of fairness, data security and legal compliance. Conclusion Legal evolutions and tech advancements in 2025 have triggered a fundamental paradigm shift in corporate internal investigations. • Regulatory expansion via the New PRC Anti-Unfair Competition Law demands holistic, extraterrito - rial investigative perspectives to identify emerging risks such as bribe acceptance and data abuse. Meanwhile, non-compete reforms refocus inves -
125 CHAMBERS.COM
Powered by FlippingBook