A group of seven Republican senators in the United States has urged the Department of Commerce to initiate an investigation into the activities of the Chinese company DeepSeek, which has developed an open-source artificial intelligence model, R1. Lawmakers are concerned that this technology could be used to support Chinese military structures and to steal confidential data from American users.
This is reported by Finway
Threats Associated with DeepSeek’s Open Source
The senators point out that the open-source nature of the model creates risks of misuse, including the potential for generating harmful content, such as instructions for making biological weapons. The R1 model has also demonstrated the ability to create social media materials that could provoke self-harm among teenagers, raising serious concerns among experts and politicians.
“DeepSeek has a history of collaborating with Chinese military and intelligence agencies and is likely to continue doing so.”
Furthermore, the model was released without proper security testing, which, according to the senators, increases the risk of leakage or malicious exploitation of confidential information.
Expert Opinions and Consequences of a Potential Ban
Chris Anderson, CEO of ByteNova AI, emphasized that open access to the model’s weights creates additional risks, such as remote code execution, data theft, or exploitation. He stated that when there is a lack of transparency regarding the origins and controls, companies may unwittingly jeopardize their own data or facilitate the realization of hostile objectives.
The senators stress the need to enhance the security of such technologies to prevent the leakage of confidential information or malicious exploitation. At the same time, they acknowledge the importance of developing artificial intelligence for businesses and consumers in the context of global technological competition with China.
Previously, the US Congress had already considered the possibility of banning DeepSeek due to concerns that the Chinese side could gain a strategic advantage in the field of artificial intelligence. Some experts support the idea of restrictions in the public sector; however, they warn that total bans could lead to the emergence of an oligopoly among American AI companies, slow down innovation, and complicate the process of technology decentralization.
It is worth noting that Dario Amodei, CEO of Anthropic, described DeepSeek as the “worst” in a data security test among AI models earlier this year.