The Pentagon has blocked chinese DeepSeek for its employees because it leaks data

The US Department of Defense is taking emergency measures to block the Chinese AI chat service DeepSeek after it became known that employees of the department were connecting to the service from work computers. This incident has raised concerns about national security, as the platform runs on Chinese servers and is subject to local legislation that requires cooperation with intelligence agencies.

According to Bloomberg, Pentagon employees used DeepSeek for at least two days, despite the possible risks of data leakage. In this regard, the US military department has begun the process of blocking the service, but access to it remains on part of the internal network.

The popularity of the Chinese chat service continues to grow, which has already led to the appearance of DeepSeek in the top downloads of the American Apple and Google app stores. In turn, on January 24, the US Navy officially banned its use among personnel, citing security threats and ethical concerns.

Meanwhile, OpenAI said it had evidence that its models were used to train DeepSeek. According to the US company, the Chinese startup used a “distillation” technique, in which data generated by large language models is used to create more compact analogues.

Concerns about information leaks led to Italy removing DeepSeek from app stores, accusing the service of violating privacy regulations.