SEOUL, South Korea’s National Intelligence Service (NIS) has raised concerns over the Chinese AI app DeepSeek, accusing it of “excessively” collecting personal data and posing national security risks.
The NIS issued an advisory urging government agencies to adopt stringent security measures when dealing with the app, which has drawn scrutiny for its data handling practices and inconsistent responses to culturally sensitive topics.
DeepSeek, a generative AI model developed in China, has been accused of collecting extensive user data, including chat logs, keyboard input patterns, and device-specific information.
DeepSeek Data Collection
The NIS revealed that this data is transmitted to servers operated by Chinese companies such as volceapplog.com. These practices are particularly concerning as they could potentially identify individual users and expose sensitive information.
The app’s privacy policy permits unrestricted sharing of user data with advertisers and third parties. South Korean authorities noted that under Chinese law, the government can access such data upon request, raising fears of state surveillance.
The NIS also highlighted that South Korean users’ data is stored on Chinese servers, posing a significant risk to national security.
The app also integrates with third-party services like CMPassport[.]com, potentially facilitating unauthorized data collection.
Despite these findings, some cybersecurity analysts argue that DeepSeek collects less data than its privacy policy suggests but caution that future updates could introduce more invasive features.
The intelligence agency flagged discrepancies in DeepSeek’s responses to questions on culturally sensitive topics.
For instance, when asked about the origin of kimchi a traditional Korean dish the app provided conflicting answers depending on the language used.
In Korean, it identified kimchi as a Korean dish, while in Chinese, it claimed the dish originated in China. These inconsistencies have reignited tensions over cultural heritage disputes between South Korea and China.
Additionally, DeepSeek has been accused of censoring politically sensitive topics. Questions about events like the 1989 Tiananmen Square crackdown prompt the app to avoid discussion by suggesting alternative topics.
South Korea joins Australia and Taiwan in warning against or restricting the use of DeepSeek due to privacy concerns.
Some South Korean government ministries have already blocked access to the app. These concerns reflect global concerns over Chinese internet companies’ data practices, which have previously resulted in restrictions on apps such as TikTok in various nations.
Experts in cybersecurity have also commented on DeepSeek’s privacy approach.
When questioned about South Korea’s actions against DeepSeek, a Chinese foreign ministry spokesperson stated that Beijing upholds strict laws on data privacy and security. The spokesperson denied allegations of government-mandated data collection through private companies.
As generative AI models gain popularity around the world, governments face the challenge of balancing innovation with privacy and security concerns.
The NIS plans to conduct a comprehensive review of DeepSeek’s technical framework in collaboration with relevant authorities. Meanwhile, South Korea’s actions signal heightened vigilance over foreign-developed AI technologies operating within its borders.
PCI DSS 4.0 & Supply Chain Attack Prevention – Free Webinar