Did you know Governments Have Access! The Terrifying Reality of AI Data Collection by ChatGPT and DeepSeek

 

We use a lot of online services that use our data for different purposes but most of the time, we do not know where our data goes and how it is being used. There are also a lot of government contracts involved that enable the collection of data behind the scenes. Most of the online services we use put users in close contact with intelligence agencies if the data isn't being directly harvested and this creates privacy risks. There are a lot of concerns about privacy violations on DeepSeek but this also took us to some concerning things about privacy on ChatGPT.

DeepSeek has data servers in China but most of the discussions about its privacy are recycled and everyone's repeating the same things. This shows that there are no new insights about information regarding privacy on these platforms but there are some new things that have emerged. The context of data protection law in China is nearly similar to the US one and the Article 35 of China’s data security law states that authorities are allowed to collect user data if it is concerned with national security or for investigating crimes. This shows that data can be obtained under specific circumstances in China as well as the US where government agencies can request user data when it comes to national security. But to obtain that data, there are some approval procedures and users are required to cooperate with the data collection.

The CLOUD Act (2018) in the US allows the authorities to access data which is stored in foreign countries if the servers are owned by businesses in the US. This allows authorities to collaborate with foreign authorities on sharing of data under specific conditions with cooperation. The US based companies like Meta, Google and Apple cooperate with law enforcement but people often forget how the law works in the US. China, as both regimes have an authoritative approach when it comes to data collection programs which have limited transparency and have many legal loopholes.

OpenAI’s ChatGPT depends on Microsoft’s Azure cloud infrastructure which is also used by the US government agencies. Microsoft also announced the launch of Azure Government Top Secret in August 2021 which will allow the government to handle top-secret data for the US intelligence agencies. This means that the infrastructure which is being used by ChatGPT is also being used by government agencies. Even though there are still some boundaries, they get blurred because OpenAI’s GPT-4 is integrated through Azure for the US government top secret cloud. This means that Azure cloud is being used for commercial applications as well as government functions.

On the other hand, data from DeepSeek is sent to China Mobile which is a state owned telecom company and is obligated to cooperate with law enforcement if it is a matter of national security. OpenAI says that it doesn't share user data with third parties, even if it's for marketing purposes but DeepSeek’s connection to China is a big concern due to several reasons. DeepSeek’s privacy policy says that it shares specific data to third parties but the policy doesn't provide any details and lacks transparency. This shows that both ChatGPT and DeepSeek are suspicious when it comes to protection of user data on these platforms.