Artists vs AI – quick links
If you’re a big tech sector executive, you might not be accustomed to hearing the word ‘no.’ So if nothing else, the Australian government’s announcement that it was ruling out a copyright fee carve-out for training AI would have been a refreshing novelty.
So how did we get here, and what does it all mean?
Artists vs AI: calling out carve-outs
Back in August, the Productivity Commission proposed a text and data mining (TDM) exemption to Australian copyright law, to allow AI developers to mine Australian-produced material without paying for it. In effect, tech companies wanting to train Large Language Models (LLMs) in Australia would be able to use locally-produced text and images without paying a cent to those who created the materials.
Scott Farquhar, CEO of the Tech Council of Australia, spoke out strongly in favour of the change. Globally, he wasn’t alone in calling for unfettered access to artists’ work. In May, former UK Deputy Prime Minister (and more recently Meta executive) Nick Clegg declared that asking creators for permission to train AI on their work would be impracticable – and if only Britain did this, ‘you would basically kill the AI industry in this country overnight’.
Understandably, this did not go down well with creatives. But no matter how big the Australian arts sector is worth economically, it’s dwarfed by the mind-pureeing sums of money involved in AI right now (which should be cause for alarm, given the lack of profits).
So it’s to the credit of the Australian government that it has sided fairly unambiguously with local creative industries over the wishes of the tech titans.
As I wrote on these pages back in August, the decision embodies basic principles of fairness: namely, that artists should be paid by those who profit from their work. The government has also made noises about art’s importance to Australian cultural life. All good stuff.
But will it matter?
Artists vs AI: symbols and sovereignty
In practical terms, it’s not clear how much of a difference this will make. Australian law can regulate what AI developers operating here pay to local creatives. But with most AI development and most data outside Australian jurisdiction, the effect of any such protection would seem to be limited.
Perhaps the main value in the Government’s decision is not simply the fact that it stared down the tech industry and defended the rights of creatives.
Politically, the decision would not have been too painful: for once, all sides of politics and the media were basically on board with protecting artists. What’s brave about the move is that, at a time when there is momentum gathering against any meaningful restraints on the advance of AI, the government asserted its democratic right to enforce limits on the technology.
Artists vs AI: technological determinism
There’s something comfortingly retro – analogue, even – about that.
AI is increasingly presented to us as an inevitability we have no real control over. The argument that Farquhar and Clegg make is that the tech is coming regardless of what we want or do, and if we try and set limits on how it can operate we’ll only be hurting ourselves and handing a competitive advantage to less scrupulous nations.
‘Technological determinism’ – the assumption that new technologies will change society regardless of culture, politics, or ethics – has become so deeply ingrained that we simply don’t notice it. We treat technology as if it’s not merely an agent, but the only agent, with the rest of us simply passive consumers who will have to accept whatever technology deigns to give us.
As soon as the world contains motor cars, the horse-drawn carriage is doomed; with the arrival of the internet, print media is done for. It doesn’t really matter if these predictions fail – for instance, video calling should have killed voice calls, but they didn’t, while text messages should not have killed voice calls, but they did. What matters is the idea that resistance is futile.
Artists vs AI: agency and creativity
Narratives about the inevitability of AI are not neutral observations. They serve specific interests, at the expense of others. Those who tell us AI cannot be rejected or controlled are not simply telling us how the world is. They’re actively making it be a certain way.
Those who seek to push back, meanwhile, are accused of being ‘luddites’ opposed to progress. But just as the original luddites were opposed to unemployment rather than technology itself, people hitting back at the demands of the tech industry for free training content simply want to ensure that creative work remains remunerative, respected, and, well, creative.
Something precious and distinctive is lost if we hand this category of work over to machines without minds – and a lot of decent, fulfilling ways to earn a living are lost that way, too.
In that sense, the government’s decision reasserts public agency when it comes to technology. Generative AI isn’t going away, but that does not mean we have no say over whether, when, and how it is used, any more than we have no say over where people can drive or how electricity is produced.
In practice, rejecting a TDM carveout may only provide a small amount of protection to Australian creatives. But it also sends the message that things can and should be otherwise – and that governments, for all their supposed impotence in the face of big business, are still prepared to act to bring better things about. It’s not nothing.
