The Chinese AI app DeepSeek has created a buzz in the AI world, drawing comparisons to the splash OpenAI made with ChatGPT. However, the attention surrounding the app could become a hurdle for its success in the U.S., as tech companies from nations deemed “adversary states” by the U.S. have learned from past experiences.
Though the app is still in its early stages, its potential national security risks have already come under scrutiny. These concerns are reminiscent of issues that have led to the downfall of U.S. sales for companies like Kaspersky and Huawei, as well as threats to the popular social media app TikTok.
Rep. John Moolenaar, R-Mich., chairman of the Select Committee on China, emphasized the potential risks posed by DeepSeek, stating, “The U.S. cannot allow CCP [Chinese Communist Party] models like DeepSeek to jeopardize our national security and use our technology to fuel their AI ambitions. We must take swift action to implement stronger export controls on technologies critical to DeepSeek’s AI infrastructure.”
DeepSeek made waves over the weekend by becoming the top downloaded app on Apple’s App Store in the U.S., surpassing ChatGPT. The Chinese app has also earned praise for its speed, efficiency, and advanced reasoning abilities.
In addition, DeepSeek operates on less powerful chips than its U.S. competitors. According to the company, these chips enable it to train its model for under $6 million, a tiny fraction of the budgets that Google, OpenAI, and Meta are spending on their models with top-tier processors.
If DeepSeek’s technology claims hold up to scrutiny, it could significantly shake up the AI industry. This could reduce the need for high-performance chipsets, lower power consumption, and diminish the demand for massive data centers, such as those planned by the Trump administration’s $500 billion Stargate project.
“DeepSeek raises important questions about the costs and investments required to achieve AGI outcomes and innovations,” said Jeff Le, former California deputy cabinet secretary.
“This race is not just about time; it also has energy and infrastructure consequences, particularly if it forces others to reconsider the recently announced Stargate project,” he added in an interview with TechNewsWorld.
Threats to National Security
Then there’s the national security issue that has caused trouble for companies like Huawei, Kaspersky, and more recently, TikTok.
In 2018, Huawei was a dominant player in the smartphone and telecommunications industry, even briefly surpassing Apple to become the second-largest smartphone maker worldwide. However, due to national security concerns, Huawei smartphones were banned in the U.S., and the company’s market share never fully recovered.
In 2024, the U.S. Department of Commerce’s Bureau of Industry and Security banned Kaspersky Lab from providing antivirus software and cybersecurity products or services in the U.S. or to U.S. citizens, citing national security risks.
The bureau determined that Kaspersky’s continued operations in the United States posed a national security risk due to the Russian government’s cyber capabilities and its potential influence over the company’s actions.
Then there’s TikTok, which the U.S. government wants removed from Chinese ownership over concerns that its parent company, ByteDance, could gather and share sensitive data from American users with the Chinese government.
Allie Mellen, a senior analyst at Forrester, a national market research firm based in Cambridge, Mass., argued that DeepSeek could present an even greater national security threat than TikTok. She highlighted that the app’s privacy policy explicitly states it can collect “your text or audio input, prompts, uploaded files, feedback, chat history, or other content” for training purposes.
“It also states it may share this information with law enforcement agencies, public authorities, and others at its discretion, with all collected data being stored in China,” Mellen told TechNewsWorld.
She added, “Furthermore, the type of information being submitted to DeepSeek is much more extensive. Users are uploading voice recordings, pictures, personal details, and even corporate data and intellectual property into the system.”
Gateway for Data Breaches
Rich Vibert, CEO of Metomic, a data privacy and security software company based in London, stated that the likelihood of the U.S. government banning DeepSeek depends on whether its capabilities are deemed a national security threat.
“If the tool is found to have the potential for large-scale exploitation of vulnerabilities or risks of leaking sensitive data, it’s likely that regulatory or security agencies will intervene to limit its use,” he told TechNewsWorld.
Such vulnerabilities were flagged on Monday by Kela, an Israeli threat intelligence firm. “Kela’s AI Red Team successfully bypassed the [DeepSeek] model in various scenarios, enabling it to produce harmful outputs, such as ransomware development, the creation of sensitive content, and detailed instructions on making toxins and explosive devices,” the company reported in a blog post.
“As AI technologies like DeepSeek become more advanced, the risks of failing to protect sensitive data increase significantly,” Vibert said.
He pointed out that while both DeepSeek and TikTok raise concerns over data security, the nature of their risks differs. “TikTok’s concerns center around the sheer scale of data collection and questions about where and how that data is stored,” he explained. “In contrast, DeepSeek poses a more focused risk, as it seems designed to identify and exploit vulnerabilities on a vast scale.”
DeepSeek raises national security concerns that go beyond the consumer privacy issues associated with TikTok, argued Gal Ringel, co-founder and CEO of MineOS, a data governance platform based in Tel Aviv, Israel. “It extends to the potential exposure of proprietary business information, trade secrets, and strategic corporate data,” he told TechNewsWorld.
“While TikTok sparked concerns about personal data exposure, DeepSeek’s AI tools present the same risks for sensitive corporate information,” he said. “Organizations must now take immediate action to audit and track their AI assets to avoid potential data exposure to China.”
“This isn’t just about knowing which AI tools are being used,” Ringel continued. “It’s about understanding where company data flows and ensuring strong protections are in place to prevent it from inadvertently falling into the wrong hands.”
“The parallels with TikTok are clear, but the stakes may be even higher when considering the potential exposure of business data to adversarial parties,” he added.
Disguised Protectionism
National security concerns could also be used as a cover for protectionist policies, much like how Apple was shielded from Huawei and today’s social media platforms are being safeguarded from TikTok.
“Trump is completely unpredictable, so it’s hard to say what will happen regarding a potential ban,” said Greg Sterling, co-founder of Near Media, a market research firm in San Francisco.
“I think it’s a bit early to speculate, but DeepSeek’s storage of U.S. data on Chinese servers, with full access by the Chinese government, makes it at least as much of a security risk as TikTok,” he told TechNewsWorld.
“The same reasoning could theoretically apply to any Chinese app,” he added. “So, the government needs to decide what its overarching policy is. The EU doesn’t allow EU citizens’ data to be stored on U.S. servers. The U.S. could take a similar stance with Chinese apps and potentially ban those that pose the greatest risks.”
