Australia leads "developed democracies" in the adoption of poorly thought-through, dangerous tech laws, thanks to its ban on working cryptography, rushed through in late 2018; now, with no debate or consultation, the Australian Parliament has passed a law that gives tech companies one hour to remove "violent materials" from their platforms with penalties for noncompliance of up to 10% of annual global turnover.
The law was rushed through in the wake of the Christchurch mosque shootings, after which the platforms completely failed to enforce their own policies, allowing millions of reposts of the footage streamed by the killer. This is part of a culture of negligence and willful blindness by the platforms, whose unwillingness to confront these matters is the stuff of legend and set them up for this outcome.
But if we think the platforms suck at moderation now, just wait until they can lose 10% of gross revenues for not having a sufficiently itchy trigger finger when it comes to censorship. We've already seen how platforms routinely block and censor the victims of crimes who are seeking justice. We've also seen that trolls are happy to expend the time and energy needed to master the policies of platforms and skate right up to them, while goading their opponents into crossing them so they can get them censored.
The EU was hoping to pass nearly identical legislation but failed to do so in the last Parliament, largely because its focus shifted to mandatory copyright filters, but the passage of the Australian law is sure to influence the European debate after May's EU elections. But these issues are all connected: once platforms are algorithmically filtering all user speech in the name of copyright, it's easy to just allow someone to shovel unlimited amounts of unappealable "terrorist" and "violent" content into the blacklist's hopper. And while computers are incapable of distinguishing copyright infringement from fair use -- they're even worse at distinguishing reports of hate crimes from hate crimes themselves.
At the same time, larding big platforms with public duties like these -- the sort of thing that costs tens or hundreds of millions to accomplish -- also ensures that we will never be able to cut them down to size and break up their monopolies. Once you deputize Big Tech with tasks that no small tech can perform, you also foreclose on any measure that might make Big Tech any smaller.
This is a catastrophe for the global internet. Between Australia's crypto ban, the EU Copyright Directive and now the panicked passage of a napkin-doodle into national law, we are watching the high-speed Chinafication of the western internet in realtime.
The chief executive of Atlassian, Scott Farquhar, said that no one wanted abhorrent violent material on the internet but “the legislation is flawed and will unnecessarily cost jobs and damage our tech industry”.
“The current legislation means that anyone working for a company that allows user generated content could potentially go to jail for [three] years,” he said.
“As written, that applies to news sites, social media sites, dating sites, job sites – anywhere user content could be created.”
Farquhar complained that the legislation failed to define how “expeditiously” violent material must be removed, and did not define who in a social media company could be punished.
Australia passes social media law penalising platforms for violent content [Paul Karp/The Guardian]