Government will follow TikTok-style ban if DeepSeek found to be unsafe

The Australian government would follow a TikTok-style ban on Chinese AI app DeepSeek should government security agencies find the app unsafe for use.

DeepSeek became the most downloaded app in Australia and the US on Tuesday. Picture: AFP

The Australian government would follow a TikTok-style ban on Chinese AI app DeepSeek and ban the Australian public service from downloading the app on their work phones should security agencies find the app unsafe for use, says Special Envoy for Cyber Security Andrew Charlton.

Dr Charlton said government security agencies were already in the midst of testing the relatively new app and would soon deliver a verdict on whether it could be used.

“As with all other applications reviewed by the security agencies, if they see a case to advise the government to take action, like we did for example with TikTok, then we’d follow that advice,” Dr Charlton said.

Andrew Charlton. Picture: NewsWire / Martin Ollman

Dr Charlton said while he hadn’t used the app yet, “I will not be downloading it on to my work phone, but that doesn’t mean I wouldn’t use it in a safer environment”.

The Chinese AI app has taken the nation by storm, becoming the No 1 most downloaded app in Apple’s app store.

Its widespread use has already seeped into Australian workplaces and companies were beginning to take action, which in some cases involved bans, said Netskope Threat Labs Director Ray Canzanese.

“Interest in DeepSeek is rapidly trending up, but the challenge with this or any popular or emerging generative AI app is the same as it was two years ago: the risk its misuse creates for organisations that haven’t implemented advanced data security controls.”

DeepSeek is having its ChatGPT moment, going through the same motions as OpenAI did near two years ago when several Australian companies banned its use in the workplace.

Dexus and Samsung were two of dozens of companies to do so in 2023 – with Dexus later walking back its ban.

DeepSeek has been banned for similar reasons, Mr Canzanese told The Australian, warning that companies fear staff would leak sensitive or confidential information into the AI application.

Mr Canzanese said there was potential that after two years of using tools like ChatGPT, ­Google’s Gemini and Meta’s Llama as well as internally developed AI tools at work, some employees may have let their guard down.

“I think that the risk now is probably because of (AI) familiarity; it’s more likely that people are going to hear about it and just pick it up and start playing with it,” he said.

Dr Charlton said Australian companies should watch closely over the use of DeepSeek and other AI applications in the workplace.

“We’d urge all Australian businesses to be careful about the way that they both use AI and also any risks that AI might introduce through a cyber perspective to their own business,” he said.

CheckPoint Software Technologies cyber security evangelist Ashwin Ram.

The Australian government had seen an increase in the use of AI to improve cyber attacks, misinformation campaigns, intellectual property theft and potential foreign interference, he said.

“I think no matter what AI you’re using, you have to be careful about the way you use it; you have to be careful about the output that it gives you; and you have to be careful about the information that you share with it because the privacy of that information varies across platforms,” Dr Charlton said.

CheckPoint Software cyber security evangelist Cyber Ashwin Ram told The Australian that cyber security criminals had already begun to manipulate DeepSeek.

“DeepSeek is still in its early days; however, not surprisingly, the Check Point Research team is already seeing cybercriminals looking to exploit it – mostly because it is free of charge, and some are claiming that it is also better at coding than ChatGPT,” he said.

“My advice to Australian businesses is to exercise caution – the same caveats that applied to OpenAI’s ChatGPT, Google Gemini and other GenAI platforms remain relevant: risks to data privacy, ethical and moral concerns regarding output, lack of disclosure for decision-making processes, explainability, transparency, and, of course, accountability when harm is caused.”

Sophos director and global field chief technology officer Chester Wisniewski also warned of cyber criminals manipulating DeepSeek.

“DeepSeek’s ‘open source’ nature opens it up for exploration – by both adversaries and enthusiasts. Like (Meta’s) Llama, it can be played with and largely have the guardrails removed,” he said.

“More pressing for companies, however, is that, due to its cost effectiveness, we are likely to see various products and companies adopt DeepSeek, which potentially carries significant privacy risks.”