TLDR: Hong Kong is expected to encounter significant difficulties in enforcing regulations on AI-generated content, particularly regarding mandatory labelling on social media platforms. This challenge arises despite mainland China’s recent implementation of a law requiring explicit and implicit identifiers for AI-generated content to combat misinformation and deepfakes. Experts attribute Hong Kong’s regulatory hurdles to its small, open market, which allows residents unrestricted access to global digital platforms, and the limitations of existing laws to adequately address the complexities of AI misuse.
Hong Kong is poised to confront substantial challenges in its efforts to regulate artificial intelligence (AI)-generated content, especially concerning the enforcement of mandatory labelling on social media platforms, according to recent expert analyses. This comes at a time when mainland China has already rolled out a new law designed to combat misinformation, deepfakes, and copyright infringements associated with AI.
Mainland China’s new regulation, issued in March and implemented on Monday, September 1, 2025, mandates that all AI-generated content online must carry clear identifiers. This includes both explicit markings visible to users and implicit identifiers, such as digital watermarks, embedded within the content’s metadata. Major Chinese social media platforms, including Tencent, Douyin, Kuaishou, Bilibili, and DeepSeek, have swiftly introduced new features to comply with this legislation. These platforms have also outlined penalties for non-compliance, ranging from limiting traffic and removing content to banning accounts and cancelling revenue streams.
Experts have highlighted the growing urgency for tighter AI regulation in Hong Kong, driven by a noticeable increase in deepfake scams and the improper use of AI technology. Globally, venture capital investment in AI surged by more than 12 times over the past decade, reaching US$110 billion (HK$858 billion) in 2024. Despite this rapid growth and the associated risks, very few jurisdictions, apart from the European Union and mainland China, have enacted dedicated AI legislation.
One of the primary obstacles for Hong Kong, as articulated by experts to the South China Morning Post, is the city’s comparatively small market size. This makes it challenging to ‘adopt an independent set of laws’ specifically for generative AI. Furthermore, the city’s ‘tiny, open market’ allows its residents unfettered access to a multitude of global social media and digital platforms, complicating any localized enforcement efforts.
Currently, Hong Kong relies on its existing legal framework, which encompasses data protection, intellectual property, anti-discrimination, and cybersecurity laws, to govern AI applications. However, critics argue that these existing statutes offer ‘insufficient protection for the public’ against the unique risks posed by AI, advocating for the introduction of dedicated AI legislation. For instance, while the Personal Data (Privacy) Ordinance (Cap. 486) addresses illegal data collection and use, including by AI, its scope may not fully cover emerging AI-specific issues.
The Hong Kong government has acknowledged the role of social media platform operators in establishing their own community guidelines and measures to manage inappropriate content. It has stated its commitment to ‘continue to keep in view global development trends in AI and make reference to international best practices, and adjust, refine and improve our AI governance framework as appropriate, having regard to the actual situation in Hong Kong.’ This proactive stance is further supported by a motion passed in the Legislative Council on November 21, 2024, titled ‘Formulating an AI Plus strategy,’ which urged the government to ‘study the feasibility of establishing an AI-related legal framework’ to enhance AI security and reliability.
In line with these efforts, the Hong Kong GAI Research and Development Centre, established in October 2023, has been tasked with studying the necessity of introducing specific rules and guidelines for AI. Research also points to alarming gaps in current AI usage, with 66% of users reportedly relying on AI output without verifying its accuracy, and 56% making workplace mistakes due to AI tools.
Also Read:
- China Unveils ‘AI Plus’ National Strategy and Mandates AI Content Labeling
- Denmark Pioneers New Copyright Protections for Personal Identity Against AI Deepfakes
Additionally, there are calls to refine the Copyright Ordinance to better accommodate AI-generated content. This move aims to attract international AI companies to Hong Kong and facilitate the global expansion of local AI firms, addressing concerns about whether existing intellectual property laws adequately protect content created by AI.
