With DeepSeek subverting the AI ​​industry, a group urges Australia to hug opportunities | Artificial Intelligence (AI)

A Australian company temporarily prevented employees from using the technology, and other companies scrambled to seek suggestions on their meaning of cybersecurity-and the Federal Minister urged to be cautious.

But others welcomed the arrival of DeepSeek and called on Australia to follow China 'leadership in developing strong and lower energy -intensive AI technology.

Since Chinese companies launched the R1 artificial intelligence model and publicly released their chat robots and applications, it has subverted the AI ​​industry.

DeepSeek shows that by training costs and processing required for the cost and processing required by training such as ChatGPT or Meta, the market value of several global industry leaders has declined, and their market value has declined.

Its arrival may marked the change of the new industry, but for the government and business, the effect is unclear. When the staff started to try new AI technology, the arrival of ChatGPT in 2022 surprised the government and enterprises, at least because of the arrival of DeepSeek, some people had a script.

Usual business

Telstra spokesman said that the company "has a strict process to evaluate all AI tools, functions and use cases in our business", including a approval of the list of AI tools, and guidelines on how to use them.

Currently in Telstra, DeepSeek is not allowed.

"Our preferred partner is MS Copilot. We are launching 21,000 Copilot 365 licenses to our employees."

Other companies should use DeepSeek to immediately seek advice.

Katherine Mansheng, an enforcement director of Cyber ​​CX, a Australian major network security company Cyber ​​CX, said that customers have sought suggestions for the company's security for the company.

"This is not surprising, because the whole world seems to have been in a faint fanaticism-whether it is economic and market tendency, and people with safety mirrors."

Deepseek and the government

Cyber ​​CX took unusual steps this week to quickly issue suggestions, including government departments and organizations that store sensitive information, and strongly considers the access to deepseek on work devices.

"We know, the government has no positive policy ... We have been on this road before." "We have already argued about Tiktok about Chinese surveillance cameras, and Huawei's Huawei in Huawei in the telecommunications network. We always take action before the facts, not before the facts ... especially because the threat is compromised around sensitive information, and you put any information about this AI assistant: it directly enters China.

"We think we need to act faster this time."

According to the federal AI policy implemented in September 2024, institutions must release transparent documents for their use of AI by the end of February 2025.

However, it is proven to understand who has made a specific use of DeepSeek in the federal government. The chief prosecutor decided to ban use Tiktok on government equipment and submitted to the Digital Transformation Bureau inquiring, and the agency in turn asked the Ministry of the Interior.

I asked the formal policy of the Ministry of the Interior on Thursday and did not provide a response at the time of publication.

Familiar debate ...

So far, some of Australia's reactions to Deepseek are familiar. On the question of how the Chinese government access to user data, some people have called for the prohibition of this technology-Huawei is banned from launching from NBN and Australia, and recent debates on Tiktok.

The Chinese government has strongly criticized the Australian Institute of Strategic Policy this week that Australia "cannot continue to make the current method of responding to the development of each new technology." It calls for the formulation of technical strategies covering AI, including investment sovereignty AI functions.

The industry Minister Ed Husic said on Tuesday that it was too early to decide whether it was a security risk.

Jump communication promotion

"If there is any risk of national interests, we will always maintain an open attitude and pay attention to what happened. I think that the conclusion is too early," he said. "But again, if we have to take action, then the government does this."

He emphasized that Australia is in the "final stage" of planning to react and will develop its own regulatory environment.

"The United States is marking their methods. The European Union has them. Canada will also adopt different methods. Our regional partners are also considering this." He said.

"We will coordinate our position and localize the method we must adopt. This is the method we will adopt."

DeepSeek was criticized.

... and new inspiration

It is recommended that Australia can learn from the achievements made by Deepseek.

According to reports, Deepseek led by previous hedge fund co -founder Liang Wenfeng is committed to studying AI instead of finding a profitable method to hire Chinese universities or relatively early senior graduates.

The Australian Science and Technology Council (including Microsoft, Atrasilians, Google and IBM) believes that Australia should use similar methods.

"DEEPSEEK has been adopted in less than two years to the world's leading technology. They have completed a lot of new graduates. This is Harry Godber, a model head of policy group that Australia should follow. explain.

"We conducted some of the greatest AI studies at the University of Australia. We also have amazing R and D carried out by Australian companies such as Atlassian and Canva."

Australia's newly appointed chief scientist Tony Haymet pointed out on Tuesday that Deepseek changed the dialogue around AI.

"Private funding in Shanghai, a group of talented 22 -year -old young people, did not get the best chips in the world, could not get NVIDIA chips, it seemed to create a better thing than the best company in the Western world." Essence

Husic announced last year that plans to formulate the AI ​​capability plan at the end of 2025, because the government also sought for mandatory guardrails to seek "high risk" AI.

Damian Kassabgi, chief executive of the Science and Technology Commission, said Australia will benefit from the cost and computing power efficiency of Deepseek.

However, he believes that the development of the ability plan is not fast enough, and other countries have made progress in AI competitions.

"When we look at our R & D investment in Australia, especially when commercial investment, we account for about the US R & D investment (for GDP expenditure)." "Therefore, we are not only behind."