Social media sites need to “lift their game” on election deepfakes, says expert

A cybersecurity expert has weighed into the issue of deepfakes as the Federal Election gets closer, suggesting social media companies have a larger role to play in patrolling the web for non-authentic content.

Professor Abhinav Dhall, from Monash University, phoned into Youth Jam's Newsflash podcast and told Isaac Mulcrone that from the "point of upload" technology companies should be liable.

"At the point of upload of any image or audio or video, on any social media platform, it should be checked for if it is a deepfake. So using deepfake detectors or watermarking techniques, the data should be evaluated," Professor Dhall told listeners.

As Australians decide whether to elect a Labor-led Government or Liberal/National-led, various videos and photos are popping up - and will continue to do so - that depict a candidate in ways that aren't authentic.

"People can be, simply, speaking something that doesn't really align with their ideologies. Participants can be shown to speak something which is not really what they've said (sic). It means a lot of disinformation is possible," he added.

To hear how to determine whether a piece of media is authentic or a deepfake click play below.