China’s DeepSeek AI application is gaining attention for its affordability and efficiency, but it raises concerns regarding censorship and data security. While it offers extensive search capabilities, it struggles with fundamental knowledge and avoids sensitive historical topics. Experts caution users about its reliability due to state-imposed censorship. The international community is scrutinizing its data practices, with countries investigating potential national security risks and data protection compliance, as China aims for global AI leadership by 2030.
The Global Impact of China’s DeepSeek AI Application
The Chinese AI application DeepSeek is making waves on the international stage. Known for its cost-effectiveness and efficiency, this technology raises important questions about security and the potential for censorship.
Understanding DeepSeek’s Limitations
DeepSeek claims to provide profound search results, much like a blue whale diving deep into the ocean for answers. However, it struggles with fundamental knowledge, such as identifying the current Chinese head of state. When queried about Xi Jinping, DeepSeek claims the question exceeds its AI capabilities. Instead, it suggests focusing on topics like mathematics and programming, indicating a clear limitation in its breadth of knowledge.
Inquiries about significant historical events, such as the Tiananmen Square protests of 1989, are similarly obstructed. These topics are effectively erased from the heavily monitored Chinese internet, highlighting the state’s control over information dissemination. This raises concerns as DeepSeek continues to gain traction in the tech community.
Merlin Münch from the Center for Trustworthy Artificial Intelligence warns users to approach the information provided by DeepSeek with skepticism due to evident censorship. This caution extends to other AI chatbots as well. Antonia Hmaidi from the China Research Institute Merics notes that while AI like ChatGPT may not always deliver accurate information, it does not systematically distort facts in a particular direction, unlike DeepSeek, which adheres to strict censorship regulations.
As China seeks to position itself as a global power, the implications of widespread use of DeepSeek are concerning. Hmaidi warns that if many individuals begin to rely on DeepSeek as a primary search engine, the integrity of democratic institutions could be compromised, making it difficult for people to discern genuine information from manipulated content.
Amidst these concerns, the international community is scrutinizing DeepSeek’s data protection practices. The US government is looking into potential national security risks posed by the app, while countries like Ireland and Italy are demanding clarity on user data usage. In fact, Italy has effectively prohibited the companies behind DeepSeek from utilizing collected data, initiating investigations into their practices.
Upon signing up for DeepSeek, users are informed about the data collected, which is stored on servers in China. Münch highlights that the company is transparent about its data collection methods, which include chat histories, personal information, and even keystroke analysis.
Recent reports of security breaches, including cyberattacks that hindered user registrations and a significant data leak, have further fueled apprehensions about DeepSeek. German authorities are also planning to assess the app, indicating that it may fall short of data protection laws. The Federal Office for Information Security has stated that it has not yet completed a thorough security evaluation of DeepSeek, urging users to consider that their interactions may not be private.
In the race for AI supremacy, the United States currently holds the lead, although China’s ambitions are rapidly advancing. The Chinese government aims to become a global leader in AI by 2030, supported by substantial state investments. With a significant number of patents in the field, China is aiming to leverage AI for various applications, from surveillance to crime prevention, with a reported 83% of Chinese companies utilizing generative AI technologies.