Technology reporter
Getty ImageEdgeik, Australian Minister of Science, has become the first member of the western government that raised privacy concerns about DeepSeek, which causes confusion in the market and high -tech industry.
From Huawei to Tiktok, Chinese technology is the subject of a claim that companies are linked to the Chinese state, fearing that human data will be harvested for the purpose of reporting Intelligence.
Donald Trump states that DeepSeek is a “wake -up call” for the United States, but does not suggest that it is a threat to national security.
However, Husic told ABC NEWS on Tuesday that there were many unexplored questions, including “data and privacy management.”
“I am very careful about it. I need to measure these types carefully,” he added.
Deepseek does not respond to the request of BBC comments, but UK and US users have not shown such caution so far.
DeepSeek states that the market analyst sensor tower has been downloaded 3 million since its launch.
These 80 % have come this week. In other words, it means that it was downloaded at three times the rate of rivals such as confusion.
Meanwhile, according to the White House’s press, Karoline Leavittt, US officials are questioning national security.
“I talked to this morning (the National Security Council). They are considering what (the impact of national security) is,” she said.
According to the CNBC, the US Navy has completely banned the use of DeepSeek apps, quoting “potential security and ethical concerns.”
The Navy did not respond immediately to comment requests from BBC News.
What data does DeepSeek collect?
According to DeepSeek’s unique privacy policy, a large amount of personal information collected from users is collected, and then saved in China to a “safe server”.
This includes the following:
“Technical information” chat -in the email address, phone number, date of birth, and history history when creating account users such as text and audio, “Technical Information” chat -IP address from mobile phones and operating systems. “Keystroke” pattern “.
It states that using this information to improve “safety, security, and stability” will improve DeepSeek.
After that, we share this information with other people, such as service providers, advertising partners, and corporate groups.
“There are true concerns about the technical potential of DeepSeek, especially its privacy policy,” said Lauren Hendry Parsons, Lauren Hendry Parsons.
She specifically emphasized the policy that can use the data “in accordance with the behavior other than the service and the service”.
However, although the app is harvested a lot, experts are very similar to the privacy policy that may have already agreed to rival services such as ChatGpt, Gemini, and social media platforms. He pointed out that there is.
So is that safe?
“In the case of an openly available AI model with a web or app interface that is not limited to these, the prompt or question asked by the AI can be used by the manufacturer of that model, just like the answer. EMILY TAYLOR, the highest executive officer of OXFORD INFORMATION LABS.
“So anyone who is working on confidential or national security needs to recognize those risks,” she told the BBC.
Dr. Richard Whitle of Salford University said that there were various concerns about data and privacy, but the models used in the United States also had “many concerns.”
“Consumers must always be cautious, especially in hype and fear of missing new and very popular apps,” he said.
The office of the Information Committee, a British data regulation authorities, is urging the people to recognize the rights used in the training of AI models.
Asked if the BBC News shared the Australian government’s concerns, and stated in a statement: To enable people to exercise information rights.
“We continue to be involved with stakeholders to promote effective transparency measures without avoiding action when the expectations of regulatory authorities are ignored.”
