DeepSeek's Rapid Rise: U.S.-China AI Competition Sparks National Security and Privacy Concerns

 Introduction

The rise of DeepSeek, a fast-rising Chinese AI lab, is stirring national security concerns in the U.S. Officials fear that its AI models could be exploited by the Chinese government for espionage, intellectual property theft, and influence campaigns. White House Press Secretary Karoline Leavitt announced that the National Security Council is investigating these security implications. The U.S. Navy has already banned DeepSeek over "potential security and ethical concerns."

DeepSeek's dominance in the Apple App Store signals a major shift in the AI competition between the U.S. and China. For years, experts have warned that AI leadership will determine economic and military supremacy. DeepSeek's rapid progress, despite U.S. restrictions, shows China's AI sector is advancing faster than expected.

Privacy experts debate whether DeepSeek's data collection policies are more invasive than those of American AI firms. Some argue its rise will be leveraged by U.S. tech giants to push for deregulation, mirroring the controversy surrounding the potential TikTok ban.

“There are completely valid privacy and data security concerns with DeepSeek,” says Calli Schroeder, AI and Human Rights lead at the Electronic Privacy Information Center (EPIC). “But all of those are present in U.S. AI products, too.”

Data and Privacy Concerns

DeepSeek functions like ChatGPT, utilizing vast datasets and cutting-edge processing capabilities. However, it is significantly cheaper to operate—training its R1 model cost just $6 million, far less than comparable U.S. models, according to Anthropic CEO Dario Amodei.

The company has also contributed to open-source AI, including its LLM v3, which rivals OpenAI’s GPT-4o. Critics worry that open access to such powerful technology could facilitate misuse, such as bioweapons creation, large-scale phishing attacks, and AI-generated misinformation. Yet AI leaders like Meta’s chief scientist Yann LeCun argue that open-source AI fosters greater innovation and accountability.

Data privacy remains a major concern. Critics allege that DeepSeek, like many AI firms, relies on sensitive or questionable datasets, potentially including leaked biometric data. David Sacks, President Donald Trump’s AI and crypto czar, has accused DeepSeek of training on OpenAI-generated content.

DeepSeek’s privacy policy states that it collects user inputs, IP addresses, and keystroke patterns, which it may use for training. Notably, it stores this data in “secure servers located in the People’s Republic of China.” Given AI’s role in personal and professional decision-making, this raises concerns about Chinese government access to private data. “If the servers are physically located in China, government access becomes much easier,” Schroeder warns.

Comparisons to TikTok’s data practices abound. “If Americans start using DeepSeek to manage their lives, the privacy risks will be akin to ‘TikTok on steroids,’” says Douglas Schmidt, dean of William & Mary’s School of Computing. He argues that AI tools collect far deeper insights into users’ personalities, preferences, and ambitions than a social media platform ever could.

Geopolitical Risks

DeepSeek is also intensifying fears of an AI arms race. Some experts suggest it brings China closer to developing Artificial General Intelligence (AGI)—AI that matches or surpasses human reasoning. While U.S. firms invest heavily in AI safety research, DeepSeek’s safety protocols remain opaque. Cybersecurity concerns have also surfaced, with the company recently limiting new sign-ups due to a "large-scale malicious attack."

Beyond cybersecurity, AI can shape ideology. All AI models exhibit biases based on their training data. Users have reported that DeepSeek’s R1 model censors sensitive topics like the Tiananmen Square massacre and affirms that Taiwan is part of China. This has sparked fears that the tool could be weaponized for political influence, espionage, or cyberattacks.

“These systems, if unchecked, can fuel disinformation, erode trust, and reinforce authoritarian narratives,” warns Ross Burley, co-founder of the Centre for Information Resilience.

In response, U.S. AI leaders and policymakers are calling for stronger investment in domestic AI development. Former President Trump recently declared that DeepSeek "should be a wake-up call" for American industries. David Sacks noted that DeepSeek’s success underscores the need for deregulation, arguing that AI competition will be fierce and that Trump was right to overturn Biden’s AI Executive Order.

These concerns could prompt stricter U.S. sanctions on Chinese tech firms—or even an outright ban on DeepSeek. The House Select Committee on the Chinese Communist Party has already called for tougher export controls on technologies supporting DeepSeek’s AI.

However, AI ethicists warn against overreaction. Ben Winters, director of AI and data privacy at the Consumer Federation of America, argues that U.S. firms exploit data just as aggressively. “Just because they’re doing it in America doesn’t make it better,” he says. He fears that DeepSeek’s rise will be used as justification for shielding American AI firms from regulation.

As these debates unfold, Schroeder advises users to exercise caution when engaging with AI tools. “If you’re concerned about a company’s origins,” she says, “be mindful of what you reveal about yourself and others.”

OpenAI Accuses DeepSeek of IP Theft

OpenAI, the company behind ChatGPT, claims it has proof that DeepSeek used its technology to create a competing artificial intelligence model, fueling concerns about intellectual property theft in the AI industry.

According to OpenAI, DeepSeek leveraged a technique called “distillation,” which enables smaller AI models to learn from larger ones. While common in AI development, OpenAI asserts that DeepSeek may have violated its terms of service by using this method to replicate its AI technology.

Security researchers at Microsoft discovered last fall that individuals linked to DeepSeek were harvesting vast troves of data through OpenAI’s API. Microsoft informed OpenAI, and both companies launched an investigation into possible unauthorized data extraction.

OpenAI’s spokesperson confirmed that the company is “working closely with the U.S. government to protect its advanced AI models.” The U.S. Navy has already banned DeepSeek due to concerns over sensitive data being accessed by the Chinese government.

David Sacks, an AI and cryptocurrency advisor to President Trump, suggested that DeepSeek may have illicitly copied OpenAI’s models. “There is strong evidence that DeepSeek engaged in model distillation using OpenAI’s technology,” he stated. Meanwhile, Josh Kushner, a major investor in OpenAI, criticized American tech leaders who have supported DeepSeek, highlighting the risks of U.S. customer data being transferred to China.

Despite the accusations, AI experts note that distillation is a widespread practice in the industry. “Many startups and researchers use responses from commercial AI models like ChatGPT to train their own models,” says Ritwik Gupta, a PhD candidate at UC Berkeley. This technique saves time and money by reducing the need for human feedback.

As OpenAI works to protect its intellectual property, the controversy highlights the broader challenge of regulating AI development across national boundaries. With DeepSeek’s latest AI model, R1, rivaling top U.S. models at a fraction of the cost, concerns over AI competition between China and the U.S. are only expected to intensify.