SHARE:
The reason why Americans are poking their heads over the Great Firewall isn’t any great mystery.
Some are doing it out of curiosity, downloading the AI-powered DeepSeek-R1 to test if it’s truly more efficient, accurate, and cost-effective than ChatGPT.
Some are doing it out of frustration or plain spite, going deeper into Chinese social media platforms like Xiaohongshu (known in the west as RedNote) in protest of the TikTok ban.
Whatever the reason, we’re seeing an influx of Americans throwing caution to the wind and downloading apps they haven’t fully vetted. And that could create problems at both the individual and organizational level.
Here’s what we see as the risk from Chinese apps like this DeepSeek-R1, RedNote, and other tools – and the precautions you should take before adding any unfamiliar tool to your work world.
What to Know about DeepSeek-R1
DeepSeek-R1, the large language model out of China, made news in January for blindsiding U.S. tech companies with performance that’s comparable to OpenAI’s o1 model at a fraction of the cost. DeepSeek claims to have trained their reasoning model for $6 million compared to OpenAI’s $100 million investment. This was allegedly done by extricating “large volumes of data possibly from OpenAI’s tools to help develop its technology, using a process called distillation.”
Early comparisons of DeepSeek-R1 suggest varying levels of performance. Some applaud DeepSeek’s productivity with their Mixture-of-Experts (MoE) model, which uses fewer computing resources to answer questions while still delivering reliably accurate and effective results. Additionally, they claim DeepSeek is better at rapid reasoning, though there is an inherent question of accuracy, especially when it comes to biases.
Though biases in AI systems have been recorded in these tools, most ethical organizations, once alerted to the partiality or discrimination, take steps to remedy this issue. With DeepSeek-R1, bias isn’t a bug, it’s a feature.
Journalists and researchers who have asked the AI model about the Tiananmen Square protests and massacres, Taiwan independence, or Uyghur persecution have found the AI chatbot to be evasive or simply repeat the Party line. Though you may not feel like this type of censorship impacts your business, you never know what critical information DeepSeek will omit or distort in their generated content. When business decisions are on the line, you can’t gamble with the truth.
What to Know about Xiaohongshu (aka RedNote)
As a blend of Instagram and TikTok, Xiaohongshu has offered an appealing social media alternative to many Americans perturbed by the current state of social media. Disgruntled and eager for change, people have accepted Mandarin-language terms and conditions without blinking an eye. However, many details about this platform should give people pause.
For starters, the literal translation of the platform’s name in Mandarin is Little Red Book, a direct reference to a collection of sayings by Chinese Communist leader Mao Zedong. We’re pro free speech, but we do think the allusion might imply the app is more akin to propaganda than a cross-continental cultural exchange.
Even though the reception of U.S. newcomers by Chinese users has generally been warm, the actions of RedNote’s leaders aren’t above questioning. They have engaged American influencers for sponsored content as the TikTok ban loomed in the hopes of spreading the word about their platform. Yet on the other side, they are scrambling to hire English language content moderators to limit the spread of unauthorized messages.
At minimum, this is a soft-power win for China, allowing them to expose Westerners to their way of thinking. At maximum, it’s an avenue for China to wage influence campaigns against the West, a fairly common tactic by the CCP during the 2024 election.
Senator Mark Warner of Virginia had it right when he posted “I’m concerned that Americans are flocking to a number of adversary-owned social media platforms. We still need a comprehensive and risk-based approach to assessing and mitigating the risks of foreign-owned apps.”
The General Risk When Using Chinese Apps
Though DeepSeek and Xiaohongshu are dominating the news, they aren’t the only apps or platforms posing a risk. Temu, Shein, and Lemon8 fall under the laws empowering the CCP to collect data from Chinese companies. Even the conversation about Huawei from six or more years ago indicates the skepticism of the risk of Chinese apps and technology to create a gateway for espionage.
There’s the Digital Silk Road initiative, which on the surface is pitched as a way of providing telecommunications, AI, cloud transformation, and other digital capabilities to developing nations. The Center for Information Security suggests that this program has enabled the CCP to “access new information flows, intelligence gathering opportunities, and host countries’ intellectual property (IP).” The CIS also sees Temu and Shein as means of the Chinese government to harvest data, collecting information on individual interests as well as trends on American consumer interests.
In general, the risk of Chinese apps boils down to these key concerns:
- Risk to Your Data – Companies are already sharing proprietary data and personal information with ChatGPT and generative AI, which has the potential to be misused. Directly giving that information to a tool like DeepSeek can funnel trade secrets and IP to the Chinese government and any business entities they want to benefit.
- Risk to Your Security – Downloading these apps, even as a comparison to domestic alternatives, can potentially create a backdoor into your organization. This can be exploited by Chinese government agents directly or by hackers who stumble upon a security weak point ripe for the picking.
- Risk to Your Compliance – Using apps subject to Chinese laws puts companies in a precarious position regarding GDPR, CCPA, and other data protection regulations. Chinese apps may be compelled to share data with their national authorities without notifying you.
- Risk to Your People – Whether your business or your people are using Chinese apps, their personal and professional lives become siphoned into CCP data harvesting practices. Private communications, location data, and contact networks can be used for social engineering, intelligence gathering, or even compromising your people (if their work falls under the various U.S. classification systems).
The challenge is that no matter the intention of the creators of these applications, their products are at the whim of the Chinese government, which has no qualms with pulling rank. Remember, if Alibaba founder Jack Ma was not immune from the pressures of the CCP, then no other tech CEO is going to be.
Picking the Right Solutions Requires Expertise
You probably can guess that we’d recommend against using any Chinese applications without deeply inspecting them. The potential risk is too high. However, where do you look when some of their alternatives offer appealing options? Start by working with an experienced IT services and solutions provider.
The associated risks to your data, security, compliance, and personnel demand careful evaluation. Rather than letting curiosity or frustration drive technology decisions, you need to partner with an IT services provider that:
- Understands your specific business needs
- Scrutinizes any tool long before adoption
- Prioritizes ethics and compliance as well as innovation
- Recommends secure alternatives with ongoing support
The Dexian team embodies all these values in our services. Expert guidance by itself isn’t enough– it’s essential to have the support of an organization that maintains the delicate balance between innovation and security. Our team of experts will keep you connected with IT solutions you can trust. That way, you don’t risk falling into a world of threats behind the Great Firewall.
SHARE: