DeepSeek rated too dodgy down under: Banned from Australian government devices
As American big tech companies lashed for their slow efforts to prevent harms
Australia’s Department of Home Affairs has banned the use of DeepSeek on federal government devices.
A policy issued on February 4th warns that the Chinese chat app conducts “extensive collection of data and exposure of that data to extrajudicial directions from a foreign government that conflict with Australian law.”
Most government entities have therefore been told to “Identify and remove all existing instances of DeepSeek products, applications and web services on all Australian Government systems and mobile devices” and prevent access to the company’s apps, web services, and other products.
Only agencies engaged in national security or regulatory work are allowed to access DeepSeek’s products, and even then they need special permission to do so.
The decision seems designed to avoid known issues with DeepSeek that include its collection of users’ keystrokes and shabby infosec practices that exposed chat histories and other data. And of course let’s not forget China’s national security and counterespionage laws that are widely felt to mean Chinese companies must share info with Beijing when asked to do so.
All of which add up to sound reasons to keep government users away from DeepSeek – a conclusion also reached by several US government agencies, the government of Taiwan, and Italy where the service has been banned for all users.
- Privacy Commissioner warns the ‘John Smiths’ of the world can acquire ‘digital doppelgangers’
- Microsoft tests 45% M365 price hikes in Asia-Pacific to see how much you enjoy AI
- Australia moves to drop some cryptography by 2030 – before quantum carves it up
- Australia lays fiendish tax trap for Meta – with an expensive escape hatch
Australia also gave American tech companies a kicking yesterday, in a review of 2021’s Online Safety Act.
The review finds the law hasn’t kept pace with the threats Australians face online.
“While the current approach of taking down harmful material, setting expectations for industry through the unenforceable Basic Online Safety Expectations and enforceable codes and standards has helped many, it has not been able to cope with the scale of problems in the online world,” the document states.
Slow action from Big Tech companies is one reason the Act is not protecting citizens as hoped.
“While some online services do more to help with user safety than others, initiatives often come way too late, don’t go far enough and occur only in response to huge amounts of public pressure,” the review states, before offering the following observation of Meta’s recent efforts to combat sextortion – the practice of threatening social media users who share explicit images of themselves and are then threatened with wide exposure of that material unless they make a payment.
This is terrific but it is an issue that has been with us and growing for around a decade, and its reforms still don’t cover all of its services. It shouldn’t have taken until this year, a year where the media has had an enormous focus on harms like sextortion and whether to limit online access for young people, for action to be taken.
Online platforms are also called out for knowing that “sensational and extreme content that drives attention and keeps people online”, and therefore having little reason to moderate content properly or limit time spent on their services. The Act is criticized for allowing only small fines that don’t deter online platforms, and recommends maximum penalties be raised to five percent of a global platform’s global turnover or AUD$50 million ($31 million) – and platforms slugged with whichever is the larger sum.
That’s just one of 67 recommendations, which includes a ban on apps that “nudify” images, further regulation of content depicting disordered eating, self-harm and substance use, and new approaches that make it easier to have complaints about online platforms heard.
Speaking of which, readers may recall that in 2024 Australian regulators ordered social media companies to take down a video depicting a knife attack on a Bishop that took place in a church. Most social media companies did so, but Elon Musk’s X challenged the order and labelled the regulator who issued it the “Australian censorship commissar”. Australia’s eSafety Commission dropped its court action on the matter while it awaited the delivery of the review mentioned above.
The Commission was still waiting for the review to land on January 24th, when it published a statement regarding the sentencing of a UK man who last year murdered three girls an injured sever others. The commissioner noted that the man searched for the church stabbing video minutes before he commencing his attacks. ®