Despite commissioning another process into how Australia might deal with artificial intelligence, the Labor government will soon have to stop studying and start making decisions about how (and how much) to regulate the rapidly ballooning technology.
Like many regulation questions in the tech space – see the under-16s social media ban, the long-delayed response to Peta Murphy’s gambling advertising report and questions about the news bargaining incentive – any move is likely to raise hackles of at least half, probably more, of the participants in the discussion.
The government finds itself caught at the nexus of business groups itching for productivity gains, unions fearful of wide swathes being cut through at-risk workforces, big tech companies eager for carve-outs and ordinary Australians pinballing between the massive benefits and the extraordinary risks of thinking machines, all-knowing search engines and uncanny chat bots.
A week on from Jim Chalmers’ economic reform summit we’re not much closer to knowing where the Albanese government will land on questions of whether overarching new legislation is needed.
The industry minister, Tim Ayres, is overseeing a national AI capability plan, expected by year’s end; a “gap analysis”, announced as an outcome of the reform roundtable, is more of a review of existing processes already on foot around AI in the areas of health, privacy, copyright and online safety. Other work focuses on incorporating AI into the public service, boosting investment in datacentres, training for AI skills, and working in international forums to help shape global rules frameworks.
The issue bleeds into nearly every portfolio and the government says it wants to take the time to get it right. But we know there are wide and varied opinions among Labor MPs and a mood among some for decisions to be reached about what the government will do – either rely on existing rules or set up a new AI act.
Despite diversity inside the caucus about exactly what to do, many are keen to see something happen to help Australia seize the moment and get ahead of the wave, or at least not be caught by it crashing on top of us.
Pointing to the government’s efforts to tackle the negative effects of social media, with its under-16s ban and plans to compensate media outlets, some Labor sources said that was a model case for why a fast-moving and generational technology should be appropriately regulated from the beginning – rather than trying to play catchup years later.
But, coming out of the roundtable, comments from various participants made it seem almost as if the only thing they could agree on about AI was that it presented some positives and some negatives. Chalmers gave vague responses about the meeting agreeing on making AI “a national priority and getting that right”.
Nearly every workplace in the country will have to figure out how to respond to AI. It’s not an easy task. From white-collar workers to mechanics, and publicans to public servants, industries will be reshaped and the face of work changed for ever. Getting it right, to maximise benefits and minimise negatives, is crucial. You don’t get a second chance to let the genie out of the bottle.
It’s a daunting task for any government. But the most immediate outcome from the reform roundtable was a plan to “accelerate work” on the national plan and to undertake a “gap analysis”, which Chalmers said would determine whether the government will pursue a new overarching law or rely on existing rules.
The previous industry and science minister, Ed Husic, was questioned by some for setting up several reviews into AI and not firming up a path of action; he had called for a stand-alone AI act. But as every major tech company races to set up new AI models, the government remains in review mode.
Perhaps a new standing committee of parliament, to constantly review the protean technology, would be helpful – but settling a broader government course of action is more urgent.
Ayres is said to be focused on capturing the benefits of the AI boom, harnessing the myriad opportunities and not missing the boat. His AI capability plan has three core pillars: capturing economic opportunities, sharing the benefits of the new technology and safety checks.
The gap analysis, announced after the roundtable, is a bit more fluid. It’s anyone’s guess how long it will take to decide whether a new law is needed.
But key players want change.
Questions remain about how to deal with copyright issues for the large learning models, and whether the government will side with tech giants seeking a “data mining” exemption or with arts, creative and media companies demanding compensation or licensing deals; a potential breakthrough between unions and the Tech Council, teased at the summit, is still to bear fruit.
Unions are still urging protection for workers and guarantees of benefits being shared.
The government may be stuck between seeing wider adoption as a major productivity measure, as well as an economic opportunity in datacentres and attracting tech capital, and protecting employees and the work they produce.
Where Labor lands on AI will be a defining issue for the government for years to come. Harnessing productivity gains, protecting workers and keeping a lid on a potentially dangerous technology is a hard balancing act – but it will probably take years to learn whether producing review upon review was the right and prudent path, or akin to standing on the beach debating the tidal wave about to crash.
