The Albanese Government’s latest online safety laws would expressly authorise Australia’s eSafety Commissioner to operate fake online identities and generate content using artificial intelligence.
The proposed Online Safety Amendment (Digital Duty of Care) Bill defines a “sock puppet identity” as a false or fictitious identity assumed while using an online service.
Under the government’s exposure draft, the eSafety Commissioner would be permitted to assume such identities to perform regulatory functions.
What’s more, the draft expressly authorises the Commissioner to use these identities to “generate material using an online service (including by means of artificial intelligence),” provided producing, possessing or disseminating that material is not otherwise illegal.
The independent statutory review of the Online Safety Act specifically recommended giving the regulator authority to use fake, anonymous or pseudonymous accounts to observe or interact with online services.
The government says the accounts are intended to test platforms and their algorithms, not to surveil individual Australians.
eSafety Commissioner Julie Inman Grant has said such accounts could, for example, allow the regulator to test whether platforms are actually applying age-assurance measures as claimed.
But why should an unelected government regulator require the legal power to create fake identities, interact anonymously with online services and generate AI material?
The powers come as the broader Digital Duty of Care proposal would impose obligations on online services to mitigate government-defined categories of harm. According to the Prime Minister’s announcement, failure to comply could attract penalties of up to $109.2 million.
But even if we assume the best intentions here, the issue is always bigger than what the present eSafety Commissioner intends to do.
Laws always outlive their initial architects. Powers granted for one supposedly benevolent purpose can later be exercised by different people, against different targets, under different definitions of “harm.”
At the very least, a government that wants the power to police fake accounts while simultaneously authorising its online safety regulator to operate fake accounts of its own should expect Australians to question its credibility.