The threat of ‘deepfake’ to Malaysia

LAST June, Facebook Inc founder Mark Zuckerberg stunned the world when he appeared in a less than 20 seconds video, declaring that “whoever controls the data, controls the future”.

The shock was short-lived as it turned out that the video was digitally altered — or a “deepfake” — where video or audio recordings make it appear as if a person said something which they did not.

The Zuckerberg video raises new concerns on the impact of deepfake, fake news and their implications on national security and corporations at large.

Experts told The Malaysian Reserve (TMR) that Malaysia needs to keep ahead of these challenges, especially when technological advancements through artificial intelligence (AI) are expected to grow rapidly in the coming years. Worse, the AI and deepfake softwares are easily available in the market.

“You can now produce deepfakes easily in your bedroom with some dedicated programmes and downloadable facesets. In short — anyone can generate a deepfake,” Munira Mustaffa, a US-based counter-intelligence expert told TMR.

“Samsung’s AI software, for instance, has made deepfake technology widely available to the general public,” she said.

Munira acknowledged that deepfake threat is fairly new to Malaysia as opposed to the US, but it will be an ongoing issue.

“It is controversial because its utility is associated with malicious intent, such as revenge, porn, propaganda, disinformation, misinformation and targeted hoaxes. Moreover, misinformation campaigns should be seen as a public health issue,” she said.

Munira recalled her experience in spotting a deepfake account in a LinkedIn profile recently.

The account, under the name Katie Jones, stated on her LinkedIn that she had a senior job at the Centre for Strategic and International Studies in Washington. “Jones” network of friends include other think tank members and even key US government officials.

“Katie Jones” had approached a friend of mine, Keir Giles on LinkedIn. Keir Giles is a well-known British expert on Russia defence and military, who is also a senior consulting fellow for Chatham House.

“Keir was immediately suspicious. He scanned her profile and reverse-searched her profile photo, which turned up zero result. Although he found nothing on her, it still did not quell his suspicion, so he turned to me and asked me to look into her. I have some experiences doing due diligence work and agreed to help,” she said.

Munira subsequently did her research on “Katie Jones” where she began to notice anomalies in the photo.

“I started to pick out the patterns and anomalies in every single one of them, and I went back to Katie Jones to study it further. Sure enough, they were all there,” she said.

Munira also learned that she had somehow inadvertently stumbled upon a Russian espionage trying to worm their way into LinkedIn to map out the network of security and defence professionals on the platform.

To this, she said, Malaysian government officials should be briefed about this threat and be selective on who they would admit into their professional network on social media.

“This preparedness should not be limited to government officials only, but also to political parties, private companies and the general public. You need to exercise caution and know who you’re letting in through the front door,” she said.

Another cybersecurity expert, Fong Choong Fook, said Malaysia has a lot of catching up to do with regard to deepfake and AI threats.

“We should look at how fast and advanced deepfake is progressing. It’s getting harder to identify deepfake because AI is getting smarter,” Fong said.

“Until we have a system that can match the ever-progressing AI, we are still catching up,” he added.

Full article from The Malaysian Reserve

 

You may also find other interesting LGMS News at the following links: