Elon Musk thumbs nose at eSafety commissioner amid damning report card

Elon Musk’s X is refusing to co-operate with the eSafety Commissioner on countering terror and violent extremism, prompting the online watchdog to claim the social media giant is engaging in ‘a lawfare pattern’.

X owner Elon Musk, with US President Donald Trump in the Oval Office at the White House, is refusing to engage with the eSafety Commissioner. Picture: Alex Brandon

You can now listen to The Australian's articles. Give us your feedback.

You can now listen to The Australian's articles.

Elon Musk’s X is refusing to co-operate with the eSafety Commissioner on countering terror, violent extremism and other harmful material, sparking accusations from the online watchdog that the billionaire’s social media platform is engaging in “a lawfare pattern”.

Ahead of releasing a report that includes warnings about AI-generated deepfakes and safety gaps across Meta, Google, Telegram and Reddit platforms, eSafety Commissioner Julie Inman Grant launched a defence against claims she was seeking to “constrict speech”.

As Mr Musk and other tech billionaires ingratiate themselves with US President Donald Trump, with the X and Tesla mogul emerging as a key figure in the White House, tech regulators across the globe are expecting increased pushback against efforts to lift transparency and safety standards online.

Responding to X’s recalcitrance, Ms Inman Grant on Wednesday said “we have tried in good faith to continue engaging with (X) on the investigations level”.

“We have heard the new (US) administration be vocal about AI safety … and what they consider censorship. To me, transparency really is not about constricting speech, it’s about more speech, it’s about being open and transparent so that we can hold companies accountable,” Ms Inman Grant told The Australian.

In a new report, released on Thursday, the federal government’s independent online safety regulator reveals “safety deficiencies” in Facebook, and that Telegram and Meta’s Messenger and WhatsApp platforms “did not employ measures to detect live-streamed terrorist and violent extremism (TVE) despite the fact that the 2019 Christchurch attack was live-streamed on Facebook Live”.

Information provided by tech firms, excluding X, include Google figures showing it received 258 user reports between April 2023 and February 2024 about suspected AI-generated deepfake terrorist or violent extremist material generated by Google’s own AI application, Gemini. Google confirmed 86 user reports of suspected AI-generated child sexual exploitation and abuse material.

Mark Zuckerberg, Lauren Sanchez, Jeff Bezos, Sundar Pichai and Elon Musk attend Donald Trump’s inauguration ceremony.

The transparency report follows the eSafety Commissioner issuing in March 2024 formal notices under the Online Safety Act to Google, Meta, WhatsApp, X, Telegram and Reddit, requiring the companies to report on steps taken to tackle terrorist and violent extremist material and activity on their platforms.

X, which has engaged in rolling legal battles with the eSafety Commissioner since Mr Musk took over the social media platform, rejected the notice and referred it to the Administrative Review Tribunal.

“There are four different cases that X has filed against us, including against the infringement notice we gave them for not answering our transparency notice around child sexual abuse material, and this was after their leader said that child sexual abuse was its top priority,” Ms Inman Grant said.

“They claimed that they didn’t have responsibility or liability when they became X (formerly Twitter) but the court ruled against them. They’ve now appealed that … So this is a lawfare pattern.”

The eSafety Commission last week slapped Telegram with an infringement notice of $957,780 for failing to respond to the transparency reporting request. Telegram has since provided information to eSafety.

Ms Inman Grant, who used to work at companies including Microsoft and Twitter, said online platforms had been exploited by online extremists and yet were not learning from such incidents.

“You’re introducing services that we’ve seen weaponised, we’ve seen be exploited, yet you’re still not building the safety protections,” she said.

“The capability is there. The intellectual capability is there, the financial capability is there, the technology is there. They’re creating the technology, for God’s sake, and they’re not eating their own dog food.”

Musk looks on as US President Donald Trump meets with Indian Prime Minister Narendra Modi in the Oval Office.

Ms Inman Grant said the report, which includes first-of-its-kind information from Telegram, indicates ongoing issues with Facebook and that WhatsApp is not prohibiting all organisations on parent company Meta’s dangerous organisations and individuals list.

“We are concerned about the safety deficiencies that still exist in Facebook today, with users unable to report live-streamed TVE in service if they are watching without being logged in. A user not logged in to YouTube faces the same hurdle,” she said.

Other key findings include no mechanisms for users not logged in to Facebook or YouTube to report live-streamed TVE, and that WhatsApp, Threads and Reddit on average take more than a day, 2.5 days and 1.3 days to respond to user reports of TVE.

Ms Inman Grant also raised concerns about inconsistencies around “baseline practices of how companies were detecting TVE using highly effective hash-matching technologies”.

These technologies involve unique digital signatures, known as a “hash”, of images that are compared against the signatures of other images to “find and identify copies with some implementations reporting an error rate as low as one in 50 billion”.

The report said Google reported that it “used hash-matching technology to detect ‘exact’, rather than altered, matches of TVE on YouTube and shared content in its consumer version of the Google Drive service, even though technology which can detect these altered variations is available”.

Ms Inman Grant said “this is deeply concerning when you consider in the first days following the Christchurch attack, Meta stated that over 800 different versions of the video were in circulation”.