The Australian government has continued advertising on X after its AI chatbot Grok praised Hitler and made antisemitic comments, despite earlier pausing ads on the platform after Elon Musk’s takeover.
The prime minister, Anthony Albanese, among other federal politicians, has also continued posting on X, after launching a proposed plan to combat antisemitism in Australia this week.
On Wednesday, Musk’s artificial intelligence firm xAI deleted “inappropriate” posts on X after Grok began praising Adolf Hitler, referring to itself as “MechaHitler” and making antisemitic comments in response to user queries.
The department of finance, which manages federal government advertising on social media, said its advertising was not paused when this incident occurred, or after.
“Brand safety for media platforms is assessed according to a range of criteria, including ensuring that appropriate policies and technology are in place to mitigate the risk of Australian government advertising appearing adjacent to inappropriate content,” the spokesperson said.
“The [advertising] agency has not recommended that government advertising on X should be paused at this time. Assessments of brand safety on all accessed media platforms, including X, is ongoing and subject to change.”
The government’s ad spend on X is relatively small compared with all media. Guardian Australia revealed in April that in the first year after Musk bought the platform, $2.7m was spent on advertising on X, compared with the total digital ad spending for 2022-2023 of $56.3m.
Several ministers in the Albanese government have continued to post on X in the days following Grok’s outburst.
On Friday, Albanese posted on X about trade with China, a day after launching a wide-ranging plan proposed by Australia’s antisemitism envoy.
The minister for foreign affairs, Penny Wong, posted photos on Friday from a meeting of Asean foreign ministers in Malaysia.
Axel Bruns, a professor in the Digital Media Research Centre at the Queensland University of Technology (QUT), said there was no benefit for government, business, politicians, journalists or other professionals to remain on X after the Grok incident.
“To remain on a platform whose built-in AI chatbot has explicitly adopted the Nazi alias ‘MechaHitler’ for itself is wholly inappropriate. The standard you walk past is the standard you accept,” he said.
Timothy Graham, an associate professor also at QUT, said it was contradictory for politicians to take a strong stance against antisemitism, and to be communicating on a platform “that has repeatedly failed to remove antisemitic content”.
“Not only that, but [the platform] is actually generating it,” Graham said. “Grok is a fundamental part of the architecture of X. When the core features are antisemitic, I think you’ve got a problem.”
The special envoy to combat antisemitism, Jillian Segal, told ABC’s RN Breakfast on Friday she had held meetings with a number of social media platforms, including X. “They’re very keen to ensure that hate is not associated with their platform,” she said, adding: “AI is the answer.”
“They are very focused on how they can construct their algorithms to root out hate or make sure it is swept from the platform without impeding proper free speech arguments,” Segal said.
Graham said he believed part of the reluctance to shift away from X is the entrenched role of what was formerly Twitter as a means for governments and politicians to communicate directly with the public.
“It’s like the Rome of platforms – all the roads lead to it and out of it, because it’s still inherited that power of infrastructure that it had before,” he said. “Elon Musk knew that and that’s why he bought it.”
When asked what the government would do to crack down on antisemitic content on social media on Thursday, Albanese said the platforms had a social responsibility, and said he often received antisemitic replies to his posts.
“Some of those are no doubt produced by bots, but some of those are produced by individuals as well, which is why so much of this [antisemitism] report as well goes to education.”
A spokesperson for the Albanese government told Guardian Australia that like most Australian politicians and world leaders, “the prime minister engages through social media platforms, new media and traditional media to reach as many Australians as possible”.
“All social media companies need to do more to protect the vulnerable, eliminate hate speech and act as responsible corporate citizens,” the spokesperson said.
Guardian Australia has sought comment from X.
Government advertising was previously suspended on what was then Twitter on 29 September 2022, after reports of a number of brands’ advertisements appearing next to inappropriate content, a report from the Australian National Audit Office revealed in April.
A spokesperson for the finance department said at the time it was suspended “to evaluate brand safety measures” and restored shortly after.
The best public interest journalism relies on first-hand accounts from people in the know.
If you have something to share on this subject you can contact us confidentially using the following methods.
Secure Messaging in the Guardian app
The Guardian app has a tool to send tips about stories. Messages are end to end encrypted and concealed within the routine activity that every Guardian mobile app performs. This prevents an observer from knowing that you are communicating with us at all, let alone what is being said.
If you don't already have the Guardian app, download it (iOS/Android) and go to the menu. Select ‘Secure Messaging’.
SecureDrop, instant messengers, email, telephone and post
See our guide at theguardian.com/tips for alternative methods and the pros and cons of each.
