China’s new Artificial Intelligence (AI) model, DeepSeek, went live earlier this year, shaking the tech sector, markets and national security apparatus. The Chinese open-source model shocked the industry with its suspiciously low costs and claims of high performance despite using fewer microchips.
DeepSeek’s hype is likely overblown. Elon Musk and Scale AI CEO Alexandr Wang insist it relies on at least 50,000 Nvidia chips, not 10,000 as claimed, and OpenAI believes its model was stolen to train the Chinese model. Still, the threat of an authoritarian state being at the forefront of AI development is sobering. To safeguard a free and open society, the U.S. must lead in AI development and ensure the technology’s future upholds American values of free expression and not the CCP’s censorship and propaganda.
Illustrating the sharp contrast between American-developed large language models (LLMs) and China’s DeepSeek, censorship on topics considered sensitive by the CCP was immediately obvious. One user asked, “What famous picture has a man with grocery bags in front of tanks…” DeepSeek started answering seemingly accurately, typing out, “The famous picture you’re referring to is known as ‘Tank Man’ or ‘The Unknown Rebel.’ It was taken on **June 5, 1989 during the-” While generating its response, it is abruptly cut off and replaced with, “Sorry, that’s beyond my current scope. Let’s talk about something else.” Another user, also asking about Tiananmen Square, was met with a similar non-response. In contrast, in the same chat thread, he asked, “What happened in Ohio in 1970,” and was met with a response detailing the “Kent State shootings.”
The Guardian also took a look, asking the China-based chatbot questions about several heavily censored topics in China, including asking if Taiwan is a country. It responded by falsely asserting that, “Taiwan has always been an inalienable part of China’s territory…”
Large language model AI is only going to become more important. It is rapidly and widely being adopted across sectors, being used in research, law, government, healthcare, finance, media and other industries. AI is here to stay and so if America fails to lead, the West will open itself up to widespread censorship, malign influence, and manipulation. It is not the first time the CCP has used technology to inflame tensions in American life.
Rutgers University’s Network Contagion Research Institute researchers found TikTok “significantly downplayed negative content related to China,” describing the app as, “an example of ‘persuasive technologies’ China is using to shape public opinion in the West.” We’ve seen how TikTok, which China falsely claims it does not manipulate, is being used as a thought weapon. Imagine how DeepSeek, which is blatant and open about it, will be used.
An analysis by The Wall Street Journal (WSJ) also uncovered troubling data, in this case, for the experience of American teenagers on the app. WSJ investigators created test account bots, registering as 13-year-old users, and browsed TikTok’s “For You” feed. Within hours, the app began showing “highly polarized content” relating to the Israel-Hamas war. Of the eight test accounts created, five were described as falling into a war content-related rabbit hole within the first 100 videos after the first instance of that content appeared, and two others after 250 videos, the majority of it pro-Palestinian.
While the WSJ report is not evidence that TikTok’s breakdown of pro-Israel vs pro-Palestinian content is significantly different from that of similar platforms, it does highlight the app as being, “uniquely powerful at picking up which videos get users’ attention and then feeding them the most engaging content on the topic.” The concern here is not that young teenagers are being exposed to political content generally, but that they’re being inundated with highly polarized, divisive narratives by an algorithm with opaque inner workings, subject to the whims of a revisionist dictatorship.
To win the AI Race, America must harness the innovative power of free enterprise.
While much more is needed to bolster the industry, President Donald Trump’s executive order cutting AI regulations, which will help get the government out of the way, and his announcement of a $500 billion private sector AI infrastructure investment are a good start. On the other hand, newly announced tariffs in Trump’s so-called “Liberation Day” speech are creating new uncertainty and disruption for the industry.
At the same time, the malign influence of China cannot be overcome without convincing Americans to reject deceitful, CCP-controlled platforms and embrace American-owned LLMs instead. TikTok’s influence on Gen Z is astronomical, with polls showing the app as their primary news source. And in the days leading up to TikTok’s divestiture deadline, three million Americans joined the Chinese app RedNote, translating literally to “Little Red Book,” the nickname for Quotations from Chairman Mao Tse-tung, perhaps the best known work of Chinese propaganda ever. The app is designed for China-based users and heavily censored. Despite this, many American users signed up without a second thought.
These are symptoms of a deeper issue, and those leading America’s institutions must effectively make the case for why the proliferation of platforms controlled by repressive governments are a danger to our way of life. Consistency would be a good start.
Despite many in Washington backing a much-needed TikTok divestiture proposal, several of those same politicians have gone on to soften their positions or even use the app in their political campaigns. American leadership must not only be consistent in highlighting the threat posed by these platforms, but also unwavering in championing the values of intellectual freedom. American politicians must be united on not using TikTok unless it is under new ownership.
Besides consistency from our elected leaders, the American tech industry must also commit to less censorship. Years of inconsistency relating to content moderation has sowed understandable doubt in much of the tech industry, and for the free world to win the battle for AI, this trust must be rebuilt. By upholding the values of free thought, expression, and transparency, US tech companies can work toward reestablishing confidence in their platforms.
To secure a free and open future, America must lead in artificial intelligence. AI’s prominence in our lives will only continue to grow, becoming central to how we seek information, communicate, and think. As the CCP continues its war on free minds everywhere, we must rise to the occasion to counter it. By empowering the private sector, speaking honestly about our adversaries, and being unapologetic in defense of our values, America can win the AI Race.
