Business Council urges Labor to avoid heavy handed approach to artificial intelligence
The calls for a light regulatory touch come as Labor mulls new mandatory guardrails for the “high risk” AI usage.
Australia risks stifling innovation and undermining safety if it pursues a heavy-handed approach to regulating artificial intelligence, the Business Council of Australia has warned.
In a new report, the industry lobby group, which represents some of Australia’s largest tech companies, as well as major financial institutions and professional service firms already deploying AI, called for a light regulatory touch to avoid burdening adopters with red tape.
“If Australia regulates too broadly – if we act out of fear – we risk shutting down the very kind of experimentation that could help us solve the problems we’re trying to address in the first place,” the report said.
In a significant step towards regulating the technology, Labor last year unveiled proposed mandatory guardrails for “high risk” AI usage, targeting areas such as healthcare, critical infrastructure and law enforcement.
A decision on the practical implementation of the guardrails, which rests with freshly minted Industry and Innovation Minister Tim Ayres, is to be determined. The government has canvassed several potential models for AI regulation: adopting the guardrails within existing laws or a stringent, economy-wide approach, likely requiring a new stand-alone AI law.
It is also considering a blend of the two alternatives that would establish an overarching framework to amend existing industry-specific legislation.
The BCA rejected a need to immediately press ahead with extra regulation of the technology, urging the government to first review existing laws to pinpoint any “clear and compelling” AI-related gaps before proceeding.
Only then, it said, should the government proceed with establishing the proposed legislative framework, limiting it to high-risk AI usage with significant potential for harm.
Low-risk applications, it added, should not be subject to regulatory oversight.
In adopting any new legislative framework, the BCA argued existing regulators – who already possess deep expertise in specific products, industries and professions – would be empowered to address AI-related risks within their sectors. “Framework legislation would ensure different regulators can conduct assessments consistently and apply common guardrails across diverse sectors, promoting compliance and easing implementation burdens,” it said.
The BCA warned against adoption of regulations akin to those imposed in the EU last year by setting an arbitrary limit on size and power of AI models, which have already been exceeded and thus render the technology inoperable. The curbs are having practical implications, with technology heavyweights Apple and Meta unable to release new products and features to European customers, while strict limits have been adopted in critical sectors like healthcare.
“An EU-style approach would be difficult to implement and impose overly broad and rigid obligations, slowing the adoption of AI across Australian sectors,” the BCA said.
Among its other recommendations to boost AI adoption and innovation, the BCA report called for streamlined approvals for new data centre construction and the introduction of standardised AI instruction in schools and in universities.
AUKUS review to take months
US Defense Department has extended its review of the AUKUS deal until November, raising fresh questions about Australia's defence spending commitments.
US beef decision ‘made in haste’
Cattle Australia told the government said it was “incredulous” that official documents seemed to suggest “weaker certification requirements for Mexico” than compared to Canada.