DeepSeek Triggers Global Scrutiny, Security Concerns, Federal Bans

The DeepSeek logo is seen at the offices of Chinese AI startup DeepSeek in Hangzhou, in China's eastern Zhejiang Province, on Feb. 5, 2025. STR/AFP via Getty Images

News Analysis

DeepSeek, a Chinese AI startup, has surprised markets by releasing a relatively cheap platform using slower chips yet at a performance level similar to its American counterparts. As a result, Nvidia Corp. saw its stock tumble by double digits on Jan. 27; a full recovery is yet to happen. The market reacted out of fear that the chipmaker’s dominant lead and American AI leadership had been lost.

The rise of the Chinese app has invited scrutiny of its data practices.

According to its data privacy policy, DeepSeek collects personal information—such as emails, phone numbers, birth dates, user inputs, device details, and “keystroke patterns”—and stores it on servers in China.

In addition, under Chinese counterespionage laws, Chinese companies must provide foreign user data if authorities demand it.

The possibility of data falling into the hands of the Chinese authorities triggered a wave of caution from cybersecurity experts, including Nadir Izrael, chief technology officer at U.S.-based Armis, who recommended that organizations block or limit access to the platform. Australia’s CyberCX urged nongovernment entities to do the same.

Within days, DeepSeek has been banned from government devices in Taiwan, Australia, South Korea, Canada, and the Netherlands. Italy went further by imposing an outright nationwide ban.

Lawmakers in other countries, including in the United States, are also investigating DeepSeek’s data collection practices and drafting legislation to restrict its use.

On Feb. 6, a bipartisan bill was introduced in the House to ban DeepSeek from U.S. government devices.

A week earlier, Sen. Josh Hawley (R-Mo.) proposed legislation to prevent American AI technology from advancing China’s AI capabilities by curbing U.S. investment and data flow to the communist country.

“Every dollar and gig of data that flows into Chinese AI are dollars and data that will ultimately be used against the United States,” Hawley said in his statement.

Lin Tsung-nan, a telecommunications engineering professor at National Taiwan University, told The Epoch Times that Chinese companies may be compelled to comply with state demands or even be state-owned firms disguised as private enterprises. Lin believes global distrust of DeepSeek stems from “its direct links to the Chinese Communist Party (CCP).”

DeepSeek has not responded to The Epoch Times’ requests for comment.

AI Race Accelerates

On his second day in office, President Donald Trump announced a $500 billion AI investment in the United States, calling it the “largest AI infrastructure project by far in history.” The new initiative aims to maintain America’s AI leadership, especially over China.

Trump also revoked an executive order issued by President Joe Biden, giving private companies more autonomy for AI development.

That was before DeepSeek disrupted the U.S. tech scene. Although the Chinese company released its latest model on Jan. 20, the U.S. market reacted a week later.

On Jan. 27, Trump called DeepSeek’s rise a “wake-up call” for U.S. tech companies.

The bipartisan leadership of the House China panel called for the White House to tighten export controls on advanced chips used in AI development.

“DeepSeek made extensive use of Nvidia’s H800 chip, the first chip that Nvidia designed specifically to fall outside U.S. export controls,” Rep. John Moolenaar (R-Mich.) and Rep. Raja Krishnamoorthi (D-Ill.) wrote in their letter to national security adviser Mike Waltz on Jan. 30.

Referring to China’s official name, the People’s Republic of China, the lawmakers added that “frequently updating export controls is imperative to ensure the PRC will not exploit regulatory gaps and loopholes to advance their AI ambitions.”

While tightening export controls addresses one aspect of AI security, artificial intelligence itself poses deeper ethical and societal challenges. Unlike traditional manufacturing, AI can reshape norms and values, making the ethics embedded in a given model potentially transformative for future generations.

According to cybersecurity expert Tseng Yi-Shuo at Taiwan’s Institute for National Defense and Security Research, the Chinese platform resembles ChatGPT but with a CCP-driven agenda.

Users have discovered that DeepSeek does not answer questions about politically sensitive topics, such as the Tiananmen Square Massacre, during which hundreds of pro-democracy Chinese students were killed. Such responses prompted suspicions that DeepSeek might serve as a propaganda tool rather than an unbiased platform.

Tseng told The Epoch Times that if the company conforms to the CCP’s values and rules to manipulate data and algorithms, it could be used as “a tool for cognitive warfare” to shape global opinion.