As artificial intelligence (AI) technology continues to advance, its application is expanding across various industries and companies of different sizes. However, improper or malicious use of AI has raised concerns about its safety. For instance, the first case of fraud using deepfake technology occurred in Hong Kong this year. A financial officer at a well-known multinational company's Hong Kong branch received a message from someone impersonating the CFO of the foreign headquarters and participated in an online video meeting with multiple people. The officer was eventually persuaded to transfer funds, resulting in a loss of HKD 200 million.

To address these challenges, many countries and regions are developing relevant regulations. In March 2024, the European Parliament passed the groundbreaking Artificial Intelligence Act, which began to apply in phases in August of the same year, becoming the world's first regulation specifically targeting AI. This law categorizes AI technologies based on their potential risk of harm and sets regulatory levels accordingly. The U.S. federal government is also actively promoting AI policies and regulations, encouraging all states to propose related laws and supporting measures to ensure the safe, reliable, and responsible use of AI across the nation. Additionally, China has been implementing regulations specifically aimed at "generative AI" to ensure the safety of users interacting with AI models that can generate text, images, audio, and video content.

Since AI models are primarily based on large datasets, biases and discrimination within the data can be learned and amplified by AI systems, resulting in outputs that reflect these biases, such as those related to gender, race, or age. Thus, these regulations not only aim to protect data privacy and individual security but also delve into the ethical issues that AI raises.

To help businesses use AI safely and responsibly, the International Organization for Standardization (ISO) has introduced the first international AI management system standard, ISO 42001 (Artificial Intelligence Management Systems). This standard covers how to reduce the trust gap in AI, context-based risk assessment, and best practices for critical risk and responsibility safeguards, aiming to provide clear guidance to companies. This allows them to maximize the value of AI while ensuring consumers and industries are not negatively impacted.

In Hong Kong, various companies and institutions are beginning to adapt to stricter regulations. This year, the Hong Kong Privacy Commissioner for Personal Data revised the Personal Data (Privacy) Ordinance related to AI and issued “Artificial Intelligence: The Model Personal Data Protection Framework” to address the challenges AI poses to personal data privacy. Additionally, the Hong Kong Monetary Authority issued consumer protection guidelines on the use of generative AI to all authorized institutions. The Securities and Futures Commission (SFC) of Hong Kong is also planning to issue a circular this year detailing existing rules and codes of practice concerning the use of generative AI. The Hong Kong government further announced plans to draft new cybersecurity legislation. It is clear that in the future, both AI development and usage in Hong Kong will require greater caution. 

AI is a powerful tool that can help solve complex problems and improve efficiency. However, the key lies in how responsibly this technology is used, considering both its short-term and long-term potential consequences. It's essential to strictly adhere to ethical standards and social values in the design, use, and monitoring of AI systems.

Jacky Wan

Ringus Solution Enterprise Limited

A：Units 1817-1820, Tower 1, Grand Century Place, 193 Prince Edward Rd W, Mong Kok, Hong Kong

E：info@ringus-solution.com

T：(852) 2907 6011

W：www.ringus-solution.com