eSafety commissioner orders X and Meta to remove violent videos following Sydney church stabbing

<span>Politicians and experts have raised concerns after misinformation spread on online platforms following the Bondi junction stabbing and an alleged attack in a church in Sydney.</span><span>Photograph: bombuscreative/Getty Images/iStockphoto</span>
Politicians and experts have raised concerns after misinformation spread on online platforms following the Bondi junction stabbing and an alleged attack in a church in Sydney.Photograph: bombuscreative/Getty Images/iStockphoto

Facebook’s parent company Meta and X/Twitter have been told to remove violent and distressing videos and imagery of the stabbing of a prominent Orthodox Christian leader in Sydney’s west on Monday evening.

The eSafety commissioner, Julie Inman Grant, told reporters on Tuesday that X and Meta had been issued with notices to remove material within 24 hours that depicted “gratuitous or offensive violence with a high degree of impact or detail”, with the companies facing potential fines if they fail to comply.

Related: Sydney church stabbing: police treating as terrorist attack the alleged stabbing of bishop during livestreamed mass

The notices relate to the alleged stabbing of Emmanuel at a service at the Assyrian Christ the Good Shepherd church in Wakeley just after 7pm on Monday. The service was being livestreamed and a recording of it has circulated online.

“While the majority of mainstream social media platforms have engaged with us, I am not satisfied enough is being done to protect Australians from this most extreme and gratuitous violent material circulating online,” Inman Grant said.

“That is why I am exercising my powers under the Online Safety Act to formally compel them to remove it.”

X has been contacted for comment.

After the notice was issued, Meta said it had added versions of the video to its database to ensure that it would block people attempting to upload it again in future. The company has been removing uploads as of Tuesday.

“Our priority is to protect people using our services from seeing this horrific content even if bad actors are determined to call attention to it,” a Meta spokesperson said. “We have taken steps to prevent possible copies of the incident being re-shared and are in contact with law enforcement and the eSafety commissioner’s office to provide any necessary assistance.”

Inman Grant said the level of any fines could depend on the gravity of the noncompliance, and said more removal notices to other platforms could be issued.

Notices have not been issued in relation to imagery from the fatal Bondi Junction Westfield Junction stabbing which has continued to circulate on social media since Saturday.

The prime minister, Anthony Albanese, told reporters on Tuesday he was concerned about the videos circulating online and the communications minister, Michelle Rowland, had been in contact with Inman Grant about getting videos removed.

“We remain concerned about the role of social media, including the publication of videos that can be very harmful, particularly for younger people who have access. Anyone with a phone essentially can do that,” he said.

“We continue to work with the eSafety commissioner and to use what powers are at our disposal to demand that material be taken down. I know the AFP commissioner and the security agencies are engaged in that as well.”

Those in the crowd outside the church on Monday night were being incited by inflammatory posts being spread on social media, one member of the Assyrian community, Maria, told Guardian Australia.

“They were reacting to what they were seeing on social media, there were many inflammatory posts making the rounds, people advocating for violence and the such. It was making lots of people very angry.”

Monday’s attack came just days after 40-year-old Queensland man Joel Cauchi killed five women and one man at Bondi Junction Westfield shopping centre in Sydney. In the immediate aftermath of the incident, TikTok and Meta were prepared for the possibility of the attack being streamed and shared online, as the 2019 Christchurch terror attack had been.

“Within 30 minutes of the tragic news breaking, our trust and safety team were notified and immediately activated our longstanding procedures that relate to managing these types of tragic events,” a TikTok spokesperson said.

“Moderators have been proactively removing content that breaches our community guidelines.”

Guardian Australia understands Meta began monitoring for live streams, as well as the bullying or harassment of victims and accounts praising the victims. As the victims were named, the company temporarily deactivated the accounts of those victims at the request of families.

The company also put interstitials on disturbing images to blur them in users’ feeds.

A spokesperson for the eSafety commissioner said on Monday platforms had been responsive to the office’s request to take down material that included depictions of real violence that are gratuitous or offensive and show a very high degree of impact or detail.

“In the immediate aftermath of the Bondi attack, eSafety took steps to respond to material circulating on major platforms depicting detailed scenes of death and violence given its potential to cause a significant amount of distress and harm to Australians, especially to the victims’ families and loved ones,” the spokesperson said.

Related: False claims started spreading about the Bondi Junction stabbing attack as soon as it happened

While the fears over livestreaming the Bondi attack did not eventuate, the platforms became flooded with misinformation about the event in the hours after.

As Guardian Australia reported on Monday, before Cauchi was named as the attacker, social media accounts shaped their own narrative about the tragedy’s cause without any information provided by authorities. Early on, social media accounts speculated whether the man was Muslim, with other accounts countering by speculating he was Jewish.

Despite the police naming Cauchi on Sunday, Guardian Australia has seen posts across platforms that still incorrectly name a Benjamin Cohen as the attacker. At least one post on Facebook, not mentioning that name but containing other misinformation about the attack, has been labelled as factchecked as false by Meta’s factcheck partner AAP.

The federal government is expected to introduce legislation later this year aimed requiring social media companies to toughen their policies on “content [that] is false, misleading or deceptive, and where the provision of that content on the service is reasonably likely to cause or contribute to serious harm”.

The introduction was delayed last year after initial consultation on the proposal led to claims it would stifle speech online and would not protect religious speech. Rowland said after “constructive consultation” with industry about amendments, it was considering refinements including definitions, additional transparency measures, and improving workability.

“We know that seriously harmful misinformation and disinformation spreads fast on social media,” she said “That’s why the Albanese government is committed to holding digital platforms to account for their public commitments to address this content.”