TikTok says 98pc of harmful content removed before being seen in Australia

About 98 per cent of harmful content on TikTok was automatically removed before it could be seen by users in Australia in the first quarter of the year, the platform says.

The popular video platform has fronted the Royal Commission on Antisemitism and Social Cohesion, which is examining the prevalence of online hate.

The commission has previously heard there has been a steady increase in antisemitic content on the platform since the Hamas attack on Israel on 7 October 2023, during which 1,200 Israelis were killed and hostages were taken, triggering an Israeli military offensive in Gaza.

TikTok's global head of policy, trust and safety, Zachary Hecht, told the commission all content on the platform was assessed for compliance with community guidelines at the point of upload by automatic moderation tools.

Mr Hecht, who travelled to Australia from the US, said antisemitic content was explicitly prohibited under the platform's safety and civility guidelines.

A number of examples of antisemitic content are provided to users, including Holocaust denial, hateful conspiracies against Jewish people, claims of supremacy and blaming an entire group for the actions of one person in that group.

Mr Hecht, who provided a voluntary statement to the inquiry, said the majority of content was removed before it could be viewed publicly on TikTok.

"For all of our community guidelines in [Quarter 1] in Australia, we removed 98 per cent of content proactively before a [user] report," he said.

The inquiry heard more than 336 million videos were posted to TikTok in 2025, with more than 270,000 removed under safety and civility guidelines.

The Online Hate Prevention Institute's Andre Oboler last week said TikTok had one of the best track records for removing harmful content when reported by a user, compared to other social media platforms.

An analysis by the institute showed that in a sample of more than 400 videos reported to TikTok, 64 per cent were removed.

Meta had removed 54 per cent of the 950 Facebook posts reported as offensive, and X scrubbed 24 per cent of the 1,700 posts that were flagged.

The inquiry heard in one instance, the AI moderator flagged a piece of antisemitic content as violating community guidelines and it was hidden from public view.

The creator appealed the decision and a human moderator then allowed the content to be posted.

The content was submitted for further review after coming to the attention of the advocate against antisemitism the Dor Foundation, with a human review again finding the content was not in violation of community guidelines.

Upon further escalation, Mr Hecht said the content was subsequently removed and "absolutely violates our policies".

"Our focus is on getting content moderation right at scale for billions of pieces of content.

"That being said, it does demonstrate that in some instances automated moderation might be getting the moderation correct and then human judgement can then complicate that.

"It would also be an instance where I would want to ensure that anybody who reviewed this content had targeted intervention."

The commission is also set to hear from YouTube, which is owned by Google.

Australia's public broadcasters are due to appear later in the week.

The ABC has made a submission to the inquiry.