Texas authorities are probing DeepSeek over concerns regarding its data privacy practices, potential ties to the Chinese Communist Party (CCP), and claims that its AI models compete with top industry players, according to the state’s attorney general.
Attorney General Ken Paxton issued a statement asserting that DeepSeek may be acting as a front for the CCP to compromise U.S. AI advancements and collect American user data. In line with an executive order from Governor Greg Abbott, Paxton’s office has already banned the Chinese AI platform from all state-issued devices. Additionally, Texas has informed DeepSeek that its platform may be violating the state’s data security and privacy laws.
As part of the inquiry, Texas has requested Google and Apple to provide any evaluations or documentation related to DeepSeek’s app prior to its approval in their respective app stores.
Texas is not alone in scrutinizing DeepSeek. Earlier this week, New York Governor Kathy Hochul also announced a ban, preventing the AI platform from being used on government devices and networks due to concerns over potential foreign surveillance and censorship. Lawmakers in Congress are also advocating for a federal ban on DeepSeek for similar reasons.
DeepSeek’s own privacy policy acknowledges that it collects various types of personal information, such as emails, phone numbers, birthdates, user interactions, device data, and even keystroke patterns. This information is stored on servers located in China. Under China’s counterespionage laws, Chinese firms must provide foreign user data to government authorities upon request.
Like other platforms operating in China, DeepSeek enforces strict censorship and exhibits biases aligned with the CCP’s political stance. Representative Josh Gottheimer (D-N.J.), who supports banning DeepSeek from federal government devices, stated that the CCP has shown a clear intent to exploit technology to threaten U.S. national security, spread disinformation, and collect data on American citizens.
DeepSeek is a subsidiary of High-Flyer, a Hangzhou-based company specializing in AI-driven stock market investments. The firm recently launched its DeepSeek-V3 and DeepSeek-R1 models, which reportedly rival leading AI technologies while being developed at a much lower cost.
DeepSeek claims to have trained its DeepSeek-V3 model using less-advanced chips for under $6 million—a fraction of the $100 million reportedly spent by OpenAI to train its latest ChatGPT model. This revelation caused a sharp decline in Nvidia’s stock value, with shares dropping 17 percent in a single day and nearly $600 billion being wiped from the company’s market capitalization.
Former President Donald Trump, who is leading a $500 billion private-sector initiative to bolster U.S. AI infrastructure, remarked that DeepSeek’s cost-effective approach should serve as a warning to American tech firms.
Meanwhile, OpenAI has accused DeepSeek of possibly misusing its data. The company stated that groups in China have been attempting to replicate U.S. AI models using a method called distillation, which extracts knowledge from larger models to train smaller ones. OpenAI’s terms of service explicitly prohibit the unauthorized use of its data for developing competing technologies. The company is currently reviewing evidence that DeepSeek may have engaged in this practice and plans to release more information as the investigation progresses.
0 Comments