Media: 联合早报 聯合早報 Lianhe Zaobao
OpenAI, the ower of ChatGPT, will stop providing Application Programming Interface (API) services to Hong Kong, China, and other "unsupported countries and regions" starting from July 9. The aim is to prevent companies in these countries from using ChatGPT to develop their own AI products and related traffic. This move is expected to affect companies that rely on OpenAI's API for creating large model applications ("repackaging" refers to changing the packaging to turn a use case into a small program for sale), potentially leading to their elimination. Users in Hong Kong, China, and globally may experience a change in AI user experience.
First, let's explain what an API is and its function. API stands for "Application Programming Interface." It is a set of definitions and protocols that allow different software applications to communicate and exchange data. Through an API, an application can request services or data from another application without needing to understand its internal implementation details. APIs can extract results of complex functions in a simple way without requiring knowledge of their internal workings and facilitate data exchange. For example, the Google Maps API can display maps on a website, or the X/Twitter API can retrieve tweets. Additionally, APIs can extend functionality, which is especially important in the GenAI era. For instance, a weather app can use APIs from regional observatories to obtain weather data and display it in the app without needing to collect and process the information themselves.
Impact of OpenAI terminating API services to Hong Kong and China
When OpenAI bans the use of their APIs/traffic from Hong Kong and China in July, startups relying on OpenAI's API for "repackaged" applications will be significantly impacted. APIs aid in rapid development, helping companies quickly integrate third-party services, reducing development time and costs; they also help expand markets by enabling global market entry through international APIs; and increase flexibility and foster collaboration opportunities. Since the launch of OpenAI services in November 2022, startups relying on OpenAI's API for "repackaged" applications have mushroomed, occupying various sectors. However, fierce competition, a lack of users willing to pay for continued use (favoring free services), economic downturns, and major layoffs in the tech industry have made it difficult for startups to secure venture capital funding. With stagnating revenues, GenAI-dependent startups face even greater challenges, leading to inevitable closures. Applications with advanced technology might attract tech giants for acquisition, aligning with the trend of technological transformation favoring large enterprises.
The cessation of API services to Hong Kong and China will not only result in functional limitations and incomplete data but can also be analyzed from short-, medium-, and long-term industry responses to predict its impact on users and internationally.
Short-term response
Developers and even users themselves can use VPNs or other network tools to route traffic to regions outside Hong Kong and China. However, this approach is only a temporary fix, as traces will be left, and OpenAI will likely detect and block it again.
Medium-term response
Integrate Microsoft Azure's OpenAI services. Microsoft has invested over $13 billion in OpenAI and holds a 49% stake in its for-profit organization. Given Microsoft's extensive business, it is unlikely to ban Hong Kong and China from using its services for now. However, this approach may not be sustainable for small and medium enterprises and startups with limited funds. Large multinational corporations may have greater bargaining power, allowing them to negotiate lower average costs for Microsoft services.
Long-term response
Startups and large enterprises may consider segmenting their customer bases into two groups:
1. Chinese-speaking markets in Hong Kong and China, connecting to Baidu Smart Cloud, Alibaba Cloud, Zhipu, etc.
2. Other regions in Europe, America, and English-speaking areas using OpenAI's API.
Following the announcement of OpenAI's ban in Hong Kong and China, Alibaba Cloud immediately announced that its AI development platform would provide an alternative to OpenAI API users, offering over 20 million free tokens and dedicated migration services. Simultaneously, Baidu announced the launch of their own plan offering new users free migration to "Wenxin Yiyan," with additional tokens for OpenAI API users for their flagship Ernie 3.5 model. Although the development of AI models in China has been slower, and the initial performance of Wenxin Yiyan was disappointing, the vast data scale in Greater China and other Chinese-speaking regions (especially in places like Malaysia and Singapore where many netizens favor Chinese-made apps) under centralized control means that data collection on personal information from over a billion people is broader, deeper, and more concentrated than in the English-speaking world. Catching up is only a matter of time, and focusing on Chinese language understanding may become a selling point for Chinese tech companies.
In reality, there are already examples of tech companies targeting separate customer bases: Douyin and TikTok cater to Chinese and international markets, respectively. This trend of splitting into two blocs—U.S. and China—is expected to intensify across the tech industry and global landscape.
The U.S., for national security reasons, has adopted a policy of economic and political decoupling from China. The OpenAI API ban is just one move that can impact application developers, inadvertently promoting Microsoft and Chinese tech companies, and potentially altering the global GenAI user experience. More such changes, driven by economic and technological measures, are expected post-U.S. elections, leading to significant global transformations.
OpenAI stops API services to Hong Kong & China impacting startups, enterprises and global AI user experience.
Talk to us to solve your digital challenges.
#OpenAI #AI #API #HongKong #China #GenAI #artificialintelligence #microsoft #userexperience #decoupling
