Elon Musk’s takeover of X weakened the platform’s response to antisemitism and allowed previously banned users to return, an online hate expert has told the Royal Commission into Antisemitism and Social Cohesion.
Online Hate Prevention Institute CEO Dr Andre Oboler said about 80 per cent of Twitter’s trust and safety staff were dismissed after Musk bought the company in 2022 and shifted it towards what he described as an “absolute free speech” approach. The platform was later renamed X.
Dr Oboler said the changes had made it more difficult for Australian organisations to have antisemitic and extremist content removed.
“X was generally difficult to work with, particularly from Australia,” he told the commission.
X has not responded to repeated requests from the royal commission to engage with the inquiry. Dr Oboler said it had been several years since OHPI had meaningful contact with the company.
Mr Musk has previously denied accusations he is antisemitic or racist, claims levelled in 2023 after he posted in support of a conspiracy theory involving Jews.
Oboler’s evidence showed sharp differences in the way major social media platforms responded to reports of antisemitic and extremist content.
In one OHPI sample, X removed 24 per cent of 1,700 reported posts, while Reddit removed 17 per cent of more than 1,000 posts.
TikTok removed 64 per cent of more than 400 reported videos, and Meta removed 54 per cent of 950 reported Facebook posts.
Dr Oboler said OHPI had succeeded in having harmful content removed from some platforms, while others were far less responsive.
The commission heard that online antisemitism had risen sharply following the Hamas terrorist attacks in Israel on October 7, 2023, and the subsequent war in Gaza.
Dr Oboler said OHPI had recorded a five-fold increase in antisemitic content. He said conflict involving Iran had caused a further “re-energisation” of antisemitism across social media.
Speaking outside the inquiry, he said OHPI had printed a sample of the material in “two giant folders of hate” and submitted it to the commission as part of his expert evidence.
“It’s all sorts of content, some of it quite violent,” he said.
The commission also heard that antisemitic material was often framed as political commentary, making it harder for moderation systems to identify.
Dr Oboler said LinkedIn could recognise established antisemitic tropes but often failed to act when similar material appeared in political discussion.
“The minute there’s any political discussion around it, they seem to give it a free pass, so we have content there that is extremely antisemitic, using ‘Zionist’ as a code word, but just isn’t being dealt with,” he said.
Dr Oboler said OHPI’s research had found a correlation between pro-Palestinian activism and harmful rhetoric, while stressing that the activism itself was not inherently antisemitic.
“I’m not saying that all those activities are antisemitic, far from it,” he said.
“But there is a segment of antisemitism that goes on the bandwagon with that activism, and when there’s less activism, there’s therefore obviously less opportunity for that antisemitism to rise.”
Dr Oboler appeared during the commission’s third block of public hearings, which is examining online hate, antisemitism in mainstream media and the effectiveness of platform moderation and regulation.
The ABC and SBS are also expected to appear, along with Jewish Australians who have experienced online abuse.
The hearing block has also received calls for Australia to establish permanent infrastructure to monitor hate directed at all communities, both online and offline.