Australian democracy is under threat from the rapid rise of technology such as artificial intelligence, the electoral commissioner says.
The nation's independent election overseer is also keeping a close eye on the impact of AI on the upcoming US presidential poll, where deepfake imagery has been deployed in attempts to sway voters.
Australian Electoral Commissioner Tom Rogers said his organisation had been remarkably successful in maintaining citizen trust to date.
"Maintaining this result may become increasingly complex with the rapidly expanding use of new technologies and an ever-evolving information ecosystem," he told the Accountability Round Table in Melbourne on Tuesday.
Mr Rogers said the commission was no stranger to conspiracies and disinformation, noting he received a dozen emails per week on the issue.
"This included a request for me to release all information that I have regarding the involvement of extraterrestrials in Australia's electoral system," he said.
The electoral body was mostly concerned about the increase of generative AI and how it would affect messaging at election time, Mr Rogers added.
"It will certainly enable citizens and stakeholders to generate information at a volume and velocity not seen previously," he said.
"(And) with a veracity that may make it hard to discern the truth of that information or its source."
Electoral Commissioner Tom Rogers is keeping an eye on the use of AI ahead of the next election. (Mick Tsikas/AAP PHOTOS)
Governments and regulators needed to be aware of the risks of disinformation, and the commission had implemented a number of strategies to "protect the integrity of Australian elections", Mr Rogers said.
It upped its social media presence in 2019, risking trolling and backlash to maintain a stronger presence on digital platforms.
This included social media videos of staff explaining key facts about electoral processes.
Various task forces and integrity units have also been set up across the organisation, including a disinformation register where prominent pieces of electoral disinformation are listed.
"This enables us to assure citizens that we are acting on various matters, assists us in maintaining trust, and adds a layer of simplicity and accessibility for citizens who may otherwise struggle to gain access to that level of information," Mr Rogers said.
In May, the commissioner stopped short of calling for a blanket ban on generative AI but said use of the technology during election campaigns should come with a clear warning.
Examples of AI-generated misinformation include deepfake videos pretending to deliver messages from candidates and robocalls misleading voters about how to participate in elections.
Mr Rogers said social media companies were less willing to remove disinformation than they were in the 2022 election.
He said the commission would outline an interim plan in October on how it would continue to manage AI.
"We will need to take account of what occurs in the US election ... and we may need to amend that statement before our own event," he said.