Google’s ‘bots’ can ban news

GOOGLE algorithms are automatically selecting videos from news channels to be censored from Australian viewers on the grounds they have been classified offensive, inaccurate or dangerous, a media diversity inquiry has heard.

But the tech giant has defended the move, saying its parameters were set by global “trust and safety teams” based on the best evolving advice from governments, health authorities and community standards.

The admission came during an inquiry by the Senate Environment and Communications References Committee into media diversity and a oneweek YouTube ban on Sky News and the removal of 23 clips last year related to Covid-19.

Answering a question from Senator Kim Carr, Google conceded it removed the videos after Google’s algorithm determined the footage – which the committee heard was from Sky commentators and largely related to Covid-19 – to be violations of its misinformation policies.

Google Australia and New Zealand public policy director Lucinda Longcroft said combating misinformation was a whole-of-society challenge and the tech giant took its responsibilities seriously.

The footage, the committee was told, was largely filmed before the availability of Covid-19 vaccinations and included medical experts and commentators questioning whether ivermectin and hydroxychloroquine could be a treatment for Covid-19.

Ms Longcroft said the videos were determined to be misinformation by artificial intelligence software and no personal complaint had been made to the platform.

She revealed there were 12,000 staff in the tech giant’s trust and safety teams, who were experts in the application of its community guidelines. Google removed about 5000 videos related to Covid-19 with Australian IP addresses.

But Australian Communications and Media Authority chairwoman Nerida O’Loughlin said the regulator said it did not know exactly what Sky News content Google had unilaterally decided to remove from YouTube.