AEC powerless to stop deepfake calls (21 May 2024)

Article summary: Australians face a significant threat of AI-powered disinformation in the lead-up to the next election, with the Australian Electoral Commission (AEC) expressing concern over its limited ability to combat this issue. AEC Commissioner Tom Rogers highlighted the constraints of current legislation, which allows authorised misinformation, making it lawful. This concern was raised during the first Senate inquiry hearing on the adoption of artificial intelligence, where Mr. Rogers emphasised the necessity of legislative reform to address AI-generated misinformation.

Recent incidents in other countries, such as an AI deepfake robocall impersonating US President Joe Biden during the New Hampshire presidential primary, illustrate the potential risks. Mr. Rogers noted that similar tactics in Australia, such as deepfake robocalls in marginal seats, would not breach the Electoral Act if duly authorised, underscoring the need for updated laws.

Independent Senator for Canberra David Pocock and Curtin MP Kate Chaney have taken proactive measures by writing to Trade Minister Don Farrell and Industry and Science Minister Ed Husic, urging a ban on the use of generative AI in pre-election campaigning. They stressed Australia's vulnerability to AI misuse, advocating for the establishment of a regulatory framework to prevent the deliberate spread of misinformation during election campaigns. They proposed that the recently established Artificial Intelligence Expert Group should urgently examine necessary reforms to update electoral laws.

In their letter, Pocock and Chaney called for prohibiting any AI-generated image, audio, or material during election campaigns. They argued that such legislative changes must be enacted before the next federal election, which could occur as early as August this year.

The AEC's submission to the inquiry also suggested mandatory or voluntary digital watermarking of AI-generated materials to aid in detection. Additionally, Mr. Rogers recommended a national digital literacy campaign to educate voters on identifying AI-generated misinformation, highlighting the need for comprehensive measures to safeguard Australian democracy.